chore: bump version to 0.6.1

* feat: admin Settings tab — form-based editor replacing "coming soon" stub

Section-grouped layout with collapsible headers for all ~40 ConfigStore
settings (model, session, tools, server, mcp, ratelimit, health, judge,
memory). Type-appropriate inputs: CSS toggle for bools, number with
min/max/step, select for choices, text for strings. Secret fields shown
read-only. Source badge (storage/default), amber restart indicator.

Inline save per field with dirty detection, row flash on success, reset
to default via styled confirm modal. Full WCAG keyboard accessibility
(Enter/Space on section headers, aria-labels, focus-visible). Mobile
responsive single-column at <700px. Reduced-motion safe.

* fix: Settings tab polish — help tooltips, context_window auto-detect, UX fixes

Settings UI:
- Help tooltips: ? button on ~25 settings with plain-English explanations
  and optional reference links (arXiv, Fowler, MCP spec). Click to toggle
  popover, Escape to dismiss, aria-expanded for accessibility.
- Sections start collapsed for scannable overview.
- Restart badge: hidden by default, shows when dirty, persists after save
  with amber glow. Positioned left of source badge.
- Secret row alignment fixed (transparent border matches input box model).
- Docs link in toolbar → Swagger UI Settings section.
- Number inputs: spin buttons hidden (Firefox/WebKit), empty value guard,
  numeric dirty detection (0.1 vs 0.10 no longer false positive).
- Secret reset button enabled when source=storage (clear legacy overrides).
- Space key repeat guard on section headers.
- Sidebar: sticky + max-height:100vh, no longer stretches with content.

Backend:
- context_window default changed from 131072 to 0 (auto-detect). Fallback
  lowered from 131K to 32K (realistic for local models when detection fails).
  Session normalizes 0→32768 defensively.
- Settings registry: help + reference_url fields on SettingDef, richer
  descriptions for model/session/tools/judge/memory settings.
- Schema API includes help + reference_url.
- Bootstrap system prompt: added Runtime Settings section.

Docs: tab counts updated to 13 across README, architecture, console, governance.

.dockerignore

@@ -12,3 +12,12 @@ venv/
 .mypy_cache/
 .ruff_cache/
 .hypothesis/
+deploy/
+docs/
+tests/
+sdk/
+*.md
+!README.md
+!LICENSE
+.coverage
+.swp

.env.example

@@ -1,85 +1,29 @@
 # =============================================================================
-# Turnstone Docker Compose — Environment Configuration
-# Copy to .env and fill in your values:  cp .env.example .env
+# Turnstone Environment Variables
+# Copy to .env and adjust values for your deployment
 # =============================================================================
 
-# ---------------------------------------------------------------------------
-# LLM Backend
-# ---------------------------------------------------------------------------
-# OpenAI-compatible API URL (vLLM, llama.cpp, OpenAI, etc.)
+# -- LLM Backend --------------------------------------------------------------
 LLM_BASE_URL=http://host.docker.internal:8000/v1
+OPENAI_API_KEY=sk-...
+# ANTHROPIC_API_KEY=sk-ant-...  # Set instead for Anthropic provider
+# TAVILY_API_KEY=tvly-...       # For web search fallback (local models only)
 
-# API key for the LLM backend ("dummy" for local servers without auth)
-OPENAI_API_KEY=dummy
+# -- Database (production profile) --------------------------------------------
+# DB_BACKEND=postgresql
+# POSTGRES_USER=turnstone
+# POSTGRES_PASSWORD=changeme
+# DATABASE_URL=postgresql+psycopg://turnstone:changeme@postgres:5432/turnstone
 
-# Tavily API key for web_search tool (optional)
-TAVILY_API_KEY=
+# -- Redis ---------------------------------------------------------------------
+# REDIS_PASSWORD=
+# REDIS_PORT=6379
 
-# ---------------------------------------------------------------------------
-# Redis
-# ---------------------------------------------------------------------------
-# Redis password (leave empty for no authentication)
-REDIS_PASSWORD=
+# -- Authentication ------------------------------------------------------------
+# TURNSTONE_AUTH_ENABLED=true
+# TURNSTONE_AUTH_TOKEN=your-secret-token
+# TURNSTONE_JWT_SECRET=python -c "import secrets; print(secrets.token_hex(32))"
 
-# Host port for Redis
-REDIS_PORT=6379
-
-# ---------------------------------------------------------------------------
-# Server
-# ---------------------------------------------------------------------------
-# Host port for the turnstone web UI
-SERVER_PORT=8080
-
-# Set to any non-empty value to auto-approve all tool calls
-SKIP_PERMISSIONS=
-
-# ---------------------------------------------------------------------------
-# Bridge
-# ---------------------------------------------------------------------------
-# Heartbeat TTL in seconds
-HEARTBEAT_TTL=60
-
-# Seconds to wait for external approval responses
-APPROVAL_TIMEOUT=300
-
-# ---------------------------------------------------------------------------
-# Console (Cluster Dashboard)
-# ---------------------------------------------------------------------------
-# Host port for the cluster dashboard
-CONSOLE_PORT=8090
-
-# Seconds between node polling cycles
-CONSOLE_POLL_INTERVAL=10
-
-# ---------------------------------------------------------------------------
-# Auth (optional)
-# ---------------------------------------------------------------------------
-# Set to "1" to require Bearer token authentication
-TURNSTONE_AUTH_ENABLED=
-
-# Bearer token for server/bridge/console authentication
-TURNSTONE_AUTH_TOKEN=
-
-# ---------------------------------------------------------------------------
-# Simulator (used with: docker compose --profile sim up)
-# ---------------------------------------------------------------------------
-# Number of simulated nodes
-SIM_NODES=100
-
-# Scenario: steady, burst, node_failure, directed, lifecycle
-SIM_SCENARIO=steady
-
-# Scenario duration in seconds
-SIM_DURATION=60
-
-# Messages per second (steady scenario)
-SIM_MPS=5.0
-
-# Log level
-SIM_LOG_LEVEL=INFO
-
-# Random seed for reproducibility (leave empty for random)
-SIM_SEED=
-
-# Path to write JSON metrics report (leave empty to skip)
-SIM_METRICS_FILE=
+# -- Ports ---------------------------------------------------------------------
+# SERVER_PORT=8080
+# CONSOLE_PORT=8090

.github/workflows/publish.yml

@@ -5,6 +5,7 @@ on:
     tags: ["v*"]
 
 permissions:
+  contents: write
   id-token: write
 
 jobs:
@@ -19,3 +20,10 @@ jobs:
       - run: pip install build
       - run: python -m build
       - uses: pypa/gh-action-pypi-publish@release/v1
+
+      - name: Create GitHub Release
+        uses: softprops/action-gh-release@v2
+        with:
+          generate_release_notes: true
+          draft: false
+          prerelease: ${{ contains(github.ref, '-') }}

.gitignore

@@ -17,3 +17,5 @@ venv/
 .plan.md
 .plan-*.md
 .hypothesis/
+PROGRESS.md
+.coverage

Dockerfile

@@ -25,22 +25,31 @@ FROM python:3.13-slim
 LABEL org.opencontainers.image.title="turnstone" \
       org.opencontainers.image.description="Multi-node AI orchestration platform"
 
+# System dependencies for psycopg (PostgreSQL client library)
+RUN apt-get update && apt-get install -y --no-install-recommends libpq5 \
+    && rm -rf /var/lib/apt/lists/*
+
 # Non-root user
 RUN useradd --create-home --shell /bin/bash turnstone
 
-# Install the wheel with all optional extras (redis for mq/console/sim)
+# Install the wheel with all optional extras
 COPY --from=builder /build/wheels/*.whl /tmp/wheels/
-RUN pip install --no-cache-dir "$(ls /tmp/wheels/*.whl)[mq,console,sim]" \
+RUN pip install --no-cache-dir "$(ls /tmp/wheels/*.whl)[mq,console,sim,postgres,discord]" \
     && rm -rf /tmp/wheels
 
 # Health check script (stdlib only, no pip deps needed)
 COPY docker/healthcheck.py /usr/local/bin/healthcheck.py
 
+# Entrypoint script — runs migrations before starting
+COPY docker/entrypoint.sh /usr/local/bin/entrypoint.sh
+
 # Data directory — SQLite DB is created in CWD
 WORKDIR /data
 RUN chown turnstone:turnstone /data
 
 USER turnstone
 
+ENTRYPOINT ["entrypoint.sh"]
+
 # Default command (overridden per service in compose.yaml)
 CMD ["turnstone-server", "--host", "0.0.0.0", "--port", "8080"]

QUICKSTART.md

@@ -0,0 +1,92 @@
+# Bootstrap Wizard
+
+Interactive, AI-guided setup for Turnstone deployments. Instead of manually
+editing `.env` files and reading deployment docs, the wizard walks you through
+every decision conversationally and generates all the config files for you.
+
+## Quick Start
+
+```bash
+turnstone-bootstrap
+```
+
+That's it — no flags, no arguments. The wizard prompts for everything.
+
+## How It Works
+
+1. **Pick a model** — Choose OpenAI, Anthropic, or a local/vLLM endpoint to
+   power the wizard. Local endpoints auto-detect available models.
+2. **Answer questions** — The AI walks you through deployment mode, LLM
+   provider, database, authentication, ports, and optional features.
+3. **Review generated files** — Each file is previewed before writing. You
+   confirm or reject every write.
+4. **Start the stack** — The wizard prints the exact `docker compose` command
+   and a `setup.sh` script to create your first admin user, roles, and policies.
+
+## What Gets Generated
+
+| File | Purpose |
+|------|---------|
+| `.env` | All environment variables for `compose.yaml` |
+| `setup.sh` | Post-start script: creates admin user, roles, tool policies, prompt templates via the API |
+| `docker-compose.override.yaml` | Only if customizations beyond env vars are needed |
+
+## Requirements
+
+- **Python 3.11+** with turnstone installed (`pip install turnstone`)
+- **An LLM API key** — for the wizard itself (OpenAI, Anthropic, or a local
+  model). This can differ from the LLM your deployment will use.
+- **Docker & Docker Compose** — needed to run the stack. The wizard detects
+  whether Docker is installed and gives platform-specific install instructions
+  if it's missing. You can still generate config files without Docker.
+
+## Deployment Modes
+
+The wizard supports two deployment modes:
+
+- **Single-node production** (`docker compose --profile production up`) —
+  1 server + bridge + console + PostgreSQL + Redis. Good for most use cases.
+- **Multi-node cluster** (`docker compose --profile cluster up`) —
+  10-node server/bridge fleet + PostgreSQL + Redis. For high-throughput or
+  HA deployments.
+
+## Example Session
+
+```
+$ turnstone-bootstrap
+
+ Turnstone Bootstrap Wizard  v0.5.4
+ ────────────────────────────────────────────────
+
+ Which provider for this wizard?
+   [1] OpenAI
+   [2] Anthropic
+   [3] OpenAI-compatible (local/vLLM)
+
+ > 3
+
+ Base URL [http://localhost:8000/v1]:
+ API key (press Enter for 'none'):
+
+ Querying http://localhost:8000/v1 for available models...
+ Found model: Qwen/Qwen3-32B
+
+ Connected to Qwen/Qwen3-32B. Handing off to AI assistant...
+
+> (AI walks you through the rest interactively)
+```
+
+## Tips
+
+- **Re-run safely** — running the wizard again detects your existing `.env`
+  and offers to update it rather than overwriting.
+- **Duplicate writes are skipped** — if the LLM tries to write the same file
+  twice with identical content, it's silently ignored.
+- **Type `quit` to exit** at any time during the conversation.
+- **Ctrl+C** is handled gracefully — press once to interrupt, twice to exit.
+
+## See Also
+
+- [Docker Deployment](docker.md) — manual compose setup and profiles
+- [Security](security.md) — auth architecture and token types
+- [Governance](governance.md) — roles, policies, and templates

README.md

@@ -16,16 +16,16 @@ Turnstone gives LLMs tools — shell, files, search, web, planning — and orche
 - **Interactive sessions** — terminal CLI or browser UI with parallel workstreams
 - **Queue-driven agents** — trigger workstreams via message queue, stream progress, approve or auto-approve tool use
 - **Multi-node clusters** — generic work load-balances across nodes, directed work routes to a specific server
-- **Cluster dashboard** — real-time view of all nodes, workstreams, and resource utilization
+- **Cluster dashboard** — real-time view of all nodes and workstreams, reverse proxy for server UIs
+- **Intent validation** — an LLM judge evaluates every tool call before approval, presenting risk assessments and evidence-based recommendations so users can make informed decisions instead of blindly approving raw tool calls
+- **Governance & compliance** — RBAC, tool policies, prompt templates, workstream templates, usage tracking, and append-only audit logs
 - **Cluster simulator** — test the stack at scale (up to 1000 nodes) without an LLM backend
 
-```
-External System → Message Queue → Bridge (per node) → Turnstone Server → LLM + Tools
-                                      ↓
-                                 Pub/Sub → Progress Events → External System
-                                      ↓
-                                 turnstone-console → Cluster Dashboard (browser)
-```
+Works with any OpenAI-compatible API (vLLM, llama.cpp, NVIDIA NIM) or Anthropic's native Messages API. Supports [MCP](https://modelcontextprotocol.io/) for external tool servers with native deferred tool loading on Anthropic and OpenAI APIs (BM25 fallback for local models).
+
+<p align="center">
+  <img src="docs/diagrams/architecture-overview.svg" alt="Turnstone system architecture — data flow from clients through gateways, Redis MQ, cluster nodes, to LLM providers" width="960"/>
+</p>
 
 ## Quickstart
 
@@ -72,13 +72,20 @@ pip install turnstone[console]
 turnstone-console --redis-host localhost --port 8090
 ```
 
-Then open `http://localhost:8090` for the cluster-wide dashboard.
+Then open `http://localhost:8090` for the cluster-wide dashboard. Create workstreams from the console and interact with any node's server UI through the built-in reverse proxy — no direct server port access required.
 
 ### Docker
 
 ```bash
 cp .env.example .env  # edit LLM_BASE_URL, OPENAI_API_KEY, etc.
-docker compose up     # starts redis + server + bridge + console
+docker compose up     # starts redis + server + bridge + console (SQLite)
+```
+
+For production with PostgreSQL:
+
+```bash
+# Requires POSTGRES_PASSWORD and DB_BACKEND=postgresql in .env (or exported)
+docker compose --profile production up  # adds PostgreSQL, uses it as database
 ```
 
 Console dashboard at http://localhost:8090. See [docs/docker.md](docs/docker.md) for configuration, scaling, and profiles.
@@ -100,62 +107,9 @@ turnstone-sim --nodes 100 --scenario steady --duration 60 --mps 10
 
 See [docs/simulator.md](docs/simulator.md) for scenarios, CLI reference, and metrics.
 
-All frontends connect to any OpenAI-compatible API (vLLM, NVIDIA NIM/NGC, llama.cpp, OpenAI, etc.) and auto-detect the model.
-
 ## Architecture
 
-```
-turnstone/
-├── core/              # UI-agnostic engine
-│   ├── session.py     # ChatSession — multi-turn loop, tool dispatch, agents
-│   ├── tools.py       # Tool definitions (auto-loaded from JSON)
-│   ├── workstream.py  # WorkstreamManager — parallel independent sessions
-│   ├── mcp_client.py  # MCP client manager (external tool servers)
-│   ├── model_registry.py # ModelRegistry — named models, fallback routing, per-workstream selection
-│   ├── config.py      # Unified TOML config (~/.config/turnstone/config.toml)
-│   ├── memory.py      # SQLite persistence (memories, conversations, FTS5)
-│   ├── metrics.py     # Prometheus-compatible metrics collector
-│   ├── healthcheck.py # Backend health monitor + circuit breaker
-│   ├── ratelimit.py   # Per-IP token-bucket rate limiter
-│   ├── edit.py        # File editing (fuzzy match, indentation)
-│   ├── safety.py      # Path validation, sandbox checks
-│   ├── sandbox.py     # Command sandboxing
-│   └── web.py         # Web fetch/search helpers
-├── mq/                # Message queue integration
-│   ├── protocol.py    # Typed message dataclasses (JSON serialization)
-│   ├── broker.py      # Abstract MessageBroker + RedisBroker
-│   ├── bridge.py      # Bridge service (queue ↔ HTTP API, multi-node routing)
-│   └── client.py      # TurnstoneClient — Python API for external systems
-├── console/           # Cluster dashboard
-│   ├── collector.py   # ClusterCollector — aggregates all nodes via Redis + HTTP
-│   ├── server.py      # Dashboard HTTP server + SSE
-│   └── static/        # Cluster dashboard web UI
-├── tools/             # Tool schemas (one JSON file per tool)
-├── ui/                # Frontend assets and terminal rendering
-│   └── static/        # Web UI (HTML, CSS, JS)
-├── sim/               # Cluster simulator
-│   ├── cluster.py     # SimCluster — orchestrates N nodes + dispatchers
-│   ├── node.py        # SimNode + SimWorkstream — protocol-compatible node
-│   ├── engine.py      # LLM + tool execution simulation
-│   ├── scenario.py    # 5 workload scenarios (steady, burst, node_failure, …)
-│   ├── metrics.py     # Latency, throughput, utilization collection
-│   └── cli.py         # CLI entry point (turnstone-sim)
-├── cli.py             # Terminal frontend (+ /cluster commands for console)
-├── server.py          # Web frontend (HTTP + SSE)
-└── eval.py            # Evaluation and prompt optimization harness
-docs/
-├── architecture.md    # System architecture and threading model
-├── api-reference.md   # Web server API and SSE event reference
-├── console.md         # Cluster dashboard service (turnstone-console)
-├── docker.md          # Docker Compose deployment and configuration
-├── simulator.md       # Cluster simulator usage and scenarios
-├── tools.md           # Tool schemas, execution pipeline, approval flow
-├── eval.md            # Evaluation harness internals
-└── diagrams/          # UML architecture diagrams (PlantUML sources + PNGs)
-    └── png/           # Pre-rendered diagram images
-```
-
-### Architecture Diagrams
+### Diagrams
 
 Detailed UML diagrams are available in [`docs/diagrams/`](docs/diagrams/):
 
@@ -163,8 +117,8 @@ Detailed UML diagrams are available in [`docs/diagrams/`](docs/diagrams/):
 |---------|-------------|
 | [System Context](docs/diagrams/png/01-system-context.png) | Top-level components and external dependencies |
 | [Package Structure](docs/diagrams/png/02-package-structure.png) | Python modules and dependency graph |
-| [Core Engine Classes](docs/diagrams/png/03-core-engine-classes.png) | SessionUI protocol, ChatSession, WorkstreamManager |
-| [Conversation Turn](docs/diagrams/png/04-conversation-turn.png) | Full message lifecycle through the engine |
+| [Core Engine Classes](docs/diagrams/png/03-core-engine-classes.png) | SessionUI protocol, ChatSession, LLMProvider, WorkstreamManager |
+| [Conversation Turn](docs/diagrams/png/04-conversation-turn.png) | Full message lifecycle through the engine (provider-agnostic) |
 | [Tool Pipeline](docs/diagrams/png/05-tool-pipeline.png) | Three-phase prepare/approve/execute |
 | [MQ Protocol](docs/diagrams/png/06-mq-protocol.png) | 9 inbound + 19 outbound message types |
 | [Message Routing](docs/diagrams/png/07-message-routing.png) | Multi-node routing scenarios |
@@ -173,6 +127,48 @@ Detailed UML diagrams are available in [`docs/diagrams/`](docs/diagrams/):
 | [Simulator](docs/diagrams/png/10-simulator-architecture.png) | SimCluster, dispatchers, scenarios |
 | [Console Data Flow](docs/diagrams/png/11-console-data-flow.png) | Dashboard data collection threads |
 | [Deployment](docs/diagrams/png/12-deployment.png) | Docker Compose service topology |
+| [SDK Architecture](docs/diagrams/png/13-sdk-architecture.png) | Python + TypeScript client libraries |
+| [Storage Architecture](docs/diagrams/png/14-storage-architecture.png) | Pluggable database backends (SQLite + PostgreSQL) |
+| [Auth Architecture](docs/diagrams/png/15-auth-architecture.png) | JWT, scopes, token types, login flows |
+| [Channel Architecture](docs/diagrams/png/16-channel-architecture.png) | Discord/Slack adapter protocol and routing |
+| [Notify Flow](docs/diagrams/png/17-notify-flow.png) | Channel notification dispatch |
+| [Watch Architecture](docs/diagrams/png/18-watch-architecture.png) | Periodic command polling daemon |
+| [Governance Architecture](docs/diagrams/png/19-governance-architecture.png) | RBAC, policies, audit, usage enforcement flow |
+| [WS Template Architecture](docs/diagrams/png/21-ws-template-architecture.png) | Workstream template application and lifecycle |
+| [Judge Architecture](docs/diagrams/png/22-judge-architecture.png) | Intent validation two-tier evaluation pipeline |
+
+### Governance
+
+Turnstone includes a built-in governance layer for enterprise deployments — manage who can do what, which tools run unattended, and where every token goes.
+
+- **RBAC** — 15 granular permissions, 3 built-in roles (admin / operator / viewer), custom roles, privilege escalation prevention
+- **Tool policies** — glob-pattern rules (`allow` / `deny` / `ask`) with priority ordering; automate approvals or lock down dangerous tools
+- **Prompt templates** — reusable system messages with `{{variable}}` substitution and categories
+- **Usage tracking** — per-request token and tool metrics, aggregation by day / model / user, automatic 90-day pruning
+- **Audit logging** — append-only event trail for all admin mutations, IP-aware, 365-day retention
+
+All governance features are managed through the console admin panel (13 tabs) and the full REST API. Runtime settings (model, tools, rate limiting, health, judge, memory) are configurable via the admin Settings tab — no config file edits or restarts needed for most changes. See [docs/governance.md](docs/governance.md) for setup and [docs/settings.md](docs/settings.md) for the settings reference.
+
+### Intent Validation (LLM Judge)
+
+Every tool call that requires human approval is evaluated by an intent validation judge that provides a structured risk assessment alongside the approval prompt — so instead of "approve this bash command?", users see a verdict with risk level, confidence, recommendation, and reasoning.
+
+The system uses a two-tier evaluation pipeline:
+
+1. **Heuristic tier** (instant, free) — 23 pattern-based rules classify tool calls by severity. Catches destructive commands (`rm -rf /`, `DROP TABLE`), privilege escalation (`sudo`), credential access, and more. Results appear immediately.
+2. **LLM judge tier** (async) — A full LLM evaluation runs in the background with access to `read_file` and `list_directory` for evidence gathering. The judge can inspect files that a write would overwrite, check directory contents before a delete, and cite specific evidence in its reasoning. Results update the UI progressively when ready.
+
+The judge defaults to the same model as the session (self-consistency) but can be configured to use a separate model — useful when running a small local model for tasks but wanting a commercial model for safety evaluation.
+
+```toml
+[judge]
+enabled = true           # on by default
+model = ""               # empty = same as session model
+provider = ""            # empty = same as session provider
+timeout = 60.0           # generous for local models
+```
+
+Verdicts are persisted for audit and exposed via Prometheus metrics (`turnstone_judge_verdicts_total`, `turnstone_judge_llm_latency_seconds`). See [docs/judge.md](docs/judge.md) for the full guide.
 
 ## Multi-node routing
 
@@ -197,29 +193,32 @@ Bridges BLPOP from their per-node queue (priority) then the shared queue. Direct
 
 ## Tools
 
-14 built-in tools, 2 agent tools, plus external tools via MCP:
+15 built-in tools, 2 agent tools, plus external tools via MCP:
 
 | Tool | Description | Auto-approved |
 |------|-------------|:---:|
 | `bash` | Execute shell commands | |
-| `read_file` | Read file contents | yes |
+| `read_file` | Read file contents (text or images with vision models) | yes |
 | `write_file` | Write/create files | |
 | `edit_file` | Fuzzy-match file editing | |
 | `search` | Search files by name/content | yes |
 | `math` | Sandboxed Python evaluation | |
 | `man` | Read man pages | yes |
 | `web_fetch` | Fetch URL content | |
-| `web_search` | Search via Tavily API | |
-| `remember` | Save persistent facts | yes |
-| `recall` | Search memories and history | yes |
-| `forget` | Remove a memory | yes |
+| `web_search` | Web search (provider-native or Tavily) | |
+| `memory` | Structured persistent memory (save/search/delete/list) | yes |
+| `recall` | Search conversation history | yes |
+| `notify` | Send notifications to linked channels | yes |
+| `watch` | Periodic command polling with conditions | |
 | `task` | Spawn autonomous sub-agent | |
 | `plan` | Explore codebase, write .plan.md | |
 | `mcp__*` | External tools from MCP servers | |
 
+When the total tool count exceeds a configurable threshold (default 20), MCP tools are automatically deferred using native `defer_loading` on Anthropic and OpenAI APIs, or a transparent client-side BM25 search for local models. The LLM discovers deferred tools on demand via a `tool_search` capability — no configuration needed beyond `--tool-search auto` (the default).
+
 ### MCP Tool Servers
 
-Turnstone supports the [Model Context Protocol](https://modelcontextprotocol.io/) (MCP) for connecting external tool servers. MCP tools are discovered at startup, converted to OpenAI function-calling format, and merged with built-in tools. Each MCP tool is prefixed with `mcp__{server}__{tool}` to avoid name collisions.
+Turnstone supports the [Model Context Protocol](https://modelcontextprotocol.io/) (MCP) for connecting external tool servers. MCP tools are discovered at startup, converted to OpenAI function-calling format, and merged with built-in tools. Each MCP tool is prefixed with `mcp__{server}__{tool}` to avoid name collisions. Tool lists stay fresh via push notifications (`tools.listChanged`), periodic polling for servers without push, and manual `/mcp refresh`.
 
 Configure via `config.toml` or `--mcp-config`:
 
@@ -239,30 +238,39 @@ turnstone --mcp-config ~/.config/turnstone/mcp.json
 turnstone-server --mcp-config ~/.config/turnstone/mcp.json
 ```
 
-Use `/mcp` in the REPL to list connected tools. MCP tools require user approval by default (overridden by `--skip-permissions` or UI auto-approve).
+Use `/mcp` in the REPL to list connected tools, `/mcp refresh` to re-fetch tool lists from servers. MCP tools require user approval by default (overridden by `--skip-permissions` or UI auto-approve).
 
-### Multi-Model Support
+### Multi-Model and Multi-Provider Support
 
-Turnstone supports multiple model backends per server instance. Define named models in `config.toml` and select per-workstream or switch mid-session with `/model <alias>`.
+Turnstone supports multiple model backends per server instance, including different LLM providers. `ChatSession` delegates all API communication to pluggable `LLMProvider` adapters — the internal message format stays OpenAI-like, and each provider translates at the API boundary. Define named models in `config.toml` and select per-workstream or switch mid-session with `/model <alias>`.
 
 ```toml
 [models.local]
 base_url = "http://localhost:8000/v1"
 model = "qwen3-32b"
+# provider defaults to "openai" (works with vLLM, llama.cpp, etc.)
+
+[models.claude]
+provider = "anthropic"
+api_key = "sk-ant-..."
+model = "claude-opus-4-6"
+context_window = 200000
 
 [models.openai]
 base_url = "https://api.openai.com/v1"
 api_key = "sk-..."
-model = "gpt-4o"
-context_window = 128000
+model = "gpt-5"
+context_window = 400000
 
 [model]
 default = "local"              # which model to use by default
-fallback = ["openai"]          # try these if the primary is unreachable
-agent_model = "local"          # optional: cheaper model for plan/task sub-agents
+fallback = ["claude", "openai"]  # try these if the primary is unreachable
+agent_model = "claude"         # optional: separate model for plan/task sub-agents
 ```
 
-Use `/model` to show available models, `/model openai` to switch. Workstreams created via the API accept an optional `model` parameter.
+Supported providers: `"openai"` (default -- OpenAI, vLLM, llama.cpp, any OpenAI-compatible API) and `"anthropic"` (Anthropic Messages API, requires `pip install turnstone[anthropic]`).
+
+Use `/model` to show available models, `/model claude` to switch. Workstreams created via the API accept an optional `model` parameter.
 
 ## Configuration
 
@@ -272,7 +280,7 @@ All entry points read `~/.config/turnstone/config.toml`. CLI flags override conf
 [api]
 base_url = "http://localhost:8000/v1"
 api_key = ""
-tavily_key = ""
+tavily_key = ""        # only needed for local/vLLM models without native search
 
 [model]
 name = ""              # empty = auto-detect
@@ -285,6 +293,9 @@ agent_model = ""       # model alias for plan/task sub-agents
 [tools]
 timeout = 30
 skip_permissions = false
+search = "auto"            # "auto" (enable when >threshold tools), "on", "off"
+search_threshold = 20      # min tools before tool search activates
+search_max_results = 5     # max tools returned per search query
 
 [server]
 host = "0.0.0.0"
@@ -317,8 +328,22 @@ enabled = true
 requests_per_second = 10.0
 burst = 20
 
+[database]
+backend = "sqlite"     # "sqlite" (default) or "postgresql"
+path = ".turnstone.db" # SQLite file path (relative to working directory)
+# url = "postgresql+psycopg://user:pass@host:5432/turnstone"  # PostgreSQL
+# pool_size = 5        # PostgreSQL connection pool size
+
+[judge]
+enabled = true         # intent validation for tool approvals (--no-judge to disable)
+model = ""             # empty = same as session model (self-consistency)
+provider = ""          # empty = same as session provider
+timeout = 60.0         # LLM judge timeout in seconds
+confidence_threshold = 0.7
+
 [mcp]
 config_path = ""       # path to MCP JSON config file (alternative to TOML sections)
+refresh_interval = 14400  # periodic refresh for servers without push notifications (seconds, 0 to disable)
 
 [mcp.servers.example]  # one section per MCP server
 command = "npx"
@@ -357,6 +382,9 @@ Idle workstreams are automatically cleaned up after 2 hours (configurable). In m
 - `turnstone_backend_up` — LLM backend reachability (0/1)
 - `turnstone_circuit_state` — circuit breaker state (0=closed, 1=open, 2=half_open)
 - `turnstone_workstreams_evicted_total` — workstreams auto-evicted at capacity
+- `turnstone_judge_verdicts_total{tier,risk_level}` — intent validation verdicts by tier and risk
+- `turnstone_judge_llm_latency_seconds` — LLM judge evaluation latency histogram
+- `turnstone_judge_enabled` — whether the intent validation judge is active (0/1)
 
 Per-workstream metrics are labeled by `ws_id` (bounded to 10 max workstreams).
 
@@ -373,8 +401,11 @@ Per-workstream metrics are labeled by `ws_id` (bounded to 10 max workstreams).
 ## Requirements
 
 - Python 3.11+
-- An OpenAI-compatible API endpoint ([vLLM](https://github.com/vllm-project/vllm), [NVIDIA NIM](https://build.nvidia.com/), [llama.cpp](https://github.com/ggml-org/llama.cpp), etc.)
+- An OpenAI-compatible API endpoint ([vLLM](https://github.com/vllm-project/vllm), [NVIDIA NIM](https://build.nvidia.com/), [llama.cpp](https://github.com/ggml-org/llama.cpp), etc.) or an Anthropic API key
 - Redis (for message queue bridge — `pip install turnstone[mq]`)
+- Anthropic provider (optional — `pip install turnstone[anthropic]`)
+- PostgreSQL (optional, for production — `pip install turnstone[postgres]`)
+- [Git LFS](https://git-lfs.com/) (for cloning — diagram PNGs are stored in LFS)
 
 ## License
 

compose.yaml

@@ -2,10 +2,12 @@
 # Turnstone Docker Compose Stack
 #
 # Usage:
-#   Full stack:       docker compose up
-#   With simulator:   docker compose --profile sim up
-#   Sim only:         docker compose --profile sim up redis console sim
-#   Scale bridges:    docker compose up --scale bridge=3
+#   Infra only:           docker compose up
+#   Single node:          docker compose --profile production up
+#   Production (PG):      DB_BACKEND=postgresql docker compose --profile production up
+#   10-node cluster:      docker compose --profile cluster up
+#   Cluster + DDG:        docker compose --profile ddgCluster up
+#   With simulator:       docker compose --profile sim up
 # =============================================================================
 
 name: turnstone
@@ -17,8 +19,39 @@ networks:
 volumes:
   redis-data:
   turnstone-data:
+  postgres-data:
 
 services:
+  # -------------------------------------------------------------------
+  # PostgreSQL — production database (profile: production)
+  # -------------------------------------------------------------------
+  postgres:
+    image: postgres:17-alpine
+    profiles:
+      - production
+      - cluster
+      - ddgCluster
+    environment:
+      POSTGRES_DB: turnstone
+      POSTGRES_USER: ${POSTGRES_USER:-turnstone}
+      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:?POSTGRES_PASSWORD is required for production profile}
+    volumes:
+      - postgres-data:/var/lib/postgresql/data
+    networks:
+      - turnstone-net
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER:-turnstone}"]
+      interval: 5s
+      timeout: 3s
+      retries: 5
+      start_period: 5s
+    deploy:
+      resources:
+        limits:
+          memory: 512M
+          cpus: '1.0'
+    restart: unless-stopped
+
   # -------------------------------------------------------------------
   # Redis — message broker, pub/sub, node registry
   # -------------------------------------------------------------------
@@ -57,6 +90,8 @@ services:
     build:
       context: .
       dockerfile: Dockerfile
+    profiles:
+      - production
     command:
       - sh
       - -c
@@ -66,11 +101,14 @@ services:
         --port 8080
         --base-url "$${LLM_BASE_URL}"
         --api-key "$${OPENAI_API_KEY}"
+        $${MODEL:+--model $$MODEL}
         $${SKIP_PERMISSIONS:+--skip-permissions}
+        $${MCP_CONFIG:+--mcp-config $$MCP_CONFIG}
     ports:
       - "${SERVER_PORT:-8080}:8080"
     volumes:
       - turnstone-data:/data
+      - ./docker/mcp-ddg.json:/etc/turnstone/mcp-ddg.json:ro
     environment:
       - LLM_BASE_URL=${LLM_BASE_URL:-http://host.docker.internal:8000/v1}
       - OPENAI_API_KEY=${OPENAI_API_KEY:-dummy}
@@ -78,6 +116,12 @@ services:
       - SKIP_PERMISSIONS=${SKIP_PERMISSIONS:-}
       - TURNSTONE_AUTH_ENABLED=${TURNSTONE_AUTH_ENABLED:-}
       - TURNSTONE_AUTH_TOKEN=${TURNSTONE_AUTH_TOKEN:-}
+      - TURNSTONE_JWT_SECRET=${TURNSTONE_JWT_SECRET:-}
+      - MODEL=${MODEL:-}
+      - MCP_CONFIG=${MCP_CONFIG:-}
+      - TURNSTONE_DB_BACKEND=${DB_BACKEND:-sqlite}
+      - TURNSTONE_DB_URL=${DATABASE_URL:-}
+      - TURNSTONE_NODE_ID=${TURNSTONE_NODE_ID:-}
     extra_hosts:
       - "host.docker.internal:host-gateway"
     networks:
@@ -85,6 +129,12 @@ services:
     depends_on:
       redis:
         condition: service_healthy
+      postgres:
+        condition: service_healthy
+        required: false
+      ddg-search:
+        condition: service_healthy
+        required: false
     healthcheck:
       test: ["CMD", "python", "/usr/local/bin/healthcheck.py", "http://127.0.0.1:8080/health"]
       interval: 10s
@@ -101,16 +151,19 @@ services:
     build:
       context: .
       dockerfile: Dockerfile
+    profiles:
+      - production
     command:
       - turnstone-bridge
       - --server-url=http://server:8080
       - --redis-host=redis
       - --redis-port=6379
       - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
-      - --approval-timeout=${APPROVAL_TIMEOUT:-300}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
     environment:
       - REDIS_PASSWORD=${REDIS_PASSWORD:-}
       - TURNSTONE_AUTH_TOKEN=${TURNSTONE_AUTH_TOKEN:-}
+      - TURNSTONE_JWT_SECRET=${TURNSTONE_JWT_SECRET:-}
     networks:
       - turnstone-net
     depends_on:
@@ -140,6 +193,9 @@ services:
       - REDIS_PASSWORD=${REDIS_PASSWORD:-}
       - TURNSTONE_AUTH_ENABLED=${TURNSTONE_AUTH_ENABLED:-}
       - TURNSTONE_AUTH_TOKEN=${TURNSTONE_AUTH_TOKEN:-}
+      - TURNSTONE_JWT_SECRET=${TURNSTONE_JWT_SECRET:-}
+      - TURNSTONE_DB_BACKEND=${DB_BACKEND:-sqlite}
+      - TURNSTONE_DB_URL=${DATABASE_URL:-}
     networks:
       - turnstone-net
     depends_on:
@@ -153,6 +209,78 @@ services:
       start_period: 10s
     restart: unless-stopped
 
+  # -------------------------------------------------------------------
+  # turnstone-channel — Channel gateway (Discord, Slack, etc.)
+  # Requires TURNSTONE_DISCORD_TOKEN to enable Discord adapter
+  # -------------------------------------------------------------------
+  channel:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    profiles:
+      - production
+      - cluster
+      - ddgCluster
+    command:
+      - sh
+      - -c
+      - >-
+        turnstone-channel
+        --redis-host=redis
+        --redis-port=6379
+        --http-host=0.0.0.0
+        $${TURNSTONE_DISCORD_GUILD:+--discord-guild $$TURNSTONE_DISCORD_GUILD}
+    environment:
+      - TURNSTONE_DISCORD_TOKEN=${TURNSTONE_DISCORD_TOKEN:-}
+      - TURNSTONE_DISCORD_GUILD=${TURNSTONE_DISCORD_GUILD:-0}
+      - REDIS_PASSWORD=${REDIS_PASSWORD:-}
+      - TURNSTONE_JWT_SECRET=${TURNSTONE_JWT_SECRET:-}
+      - TURNSTONE_DB_BACKEND=${DB_BACKEND:-sqlite}
+      - TURNSTONE_DB_URL=${DATABASE_URL:-}
+      - TURNSTONE_CHANNEL_ADVERTISE_URL=http://channel:8091
+    networks:
+      - turnstone-net
+    depends_on:
+      redis:
+        condition: service_healthy
+      postgres:
+        condition: service_healthy
+        required: false
+    restart: unless-stopped
+
+  # -------------------------------------------------------------------
+  # ddg-search — DuckDuckGo Search MCP server (HTTP transport)
+  # Provides web search + content fetch tools to turnstone via MCP.
+  # No API key required.
+  #
+  # Start with: MCP_CONFIG=/etc/turnstone/mcp-ddg.json \
+  #             docker compose --profile ddgCluster up
+  # -------------------------------------------------------------------
+  ddg-search:
+    image: python:3.13-slim
+    profiles:
+      - ddgCluster
+    command:
+      - sh
+      - -c
+      - >-
+        pip install --no-cache-dir duckduckgo-mcp-server &&
+        python -c "from mcp.server.transport_security import TransportSecuritySettings; import duckduckgo_mcp_server.server as s; s.safe_search=s.SafeSearchMode.OFF; s.mcp.settings.host='0.0.0.0'; s.mcp.settings.port=3000; s.mcp.settings.transport_security=TransportSecuritySettings(enable_dns_rebinding_protection=False); s.mcp.run(transport='streamable-http')"
+    networks:
+      - turnstone-net
+    healthcheck:
+      test: ["CMD-SHELL", "python -c \"import socket; s=socket.create_connection(('0.0.0.0',3000),2); s.close()\""]
+      interval: 10s
+      timeout: 5s
+      retries: 3
+      start_period: 30s
+    deploy:
+      resources:
+        limits:
+          memory: 256M
+          cpus: '0.25'
+    restart: unless-stopped
+
   # -------------------------------------------------------------------
   # turnstone-sim — Multi-node cluster simulator (no LLM needed)
   # Start with: docker compose --profile sim up
@@ -192,3 +320,224 @@ services:
       redis:
         condition: service_healthy
     restart: "no"
+
+  # ===================================================================
+  # 10-node cluster (profile: cluster)
+  #
+  # Each node is a server + bridge pair.  All share the same PostgreSQL
+  # and Redis instances.  Access via console at :8090.
+  #
+  # Start:  docker compose --profile cluster up
+  # ===================================================================
+
+  # -- cluster servers ------------------------------------------------
+
+  server-1: &cluster-server
+    build: { context: ., dockerfile: Dockerfile }
+    profiles: [cluster, ddgCluster]
+    command: &cluster-server-cmd
+      - sh
+      - -c
+      - >-
+        turnstone-server
+        --host 0.0.0.0
+        --port 8080
+        --base-url "$${LLM_BASE_URL}"
+        --api-key "$${OPENAI_API_KEY}"
+        $${MODEL:+--model $$MODEL}
+        $${SKIP_PERMISSIONS:+--skip-permissions}
+        $${MCP_CONFIG:+--mcp-config $$MCP_CONFIG}
+    volumes:
+      - turnstone-data:/data
+      - ./docker/mcp-ddg.json:/etc/turnstone/mcp-ddg.json:ro
+    environment: &cluster-server-env
+      LLM_BASE_URL: ${LLM_BASE_URL:-http://host.docker.internal:8000/v1}
+      OPENAI_API_KEY: ${OPENAI_API_KEY:-dummy}
+      TAVILY_API_KEY: ${TAVILY_API_KEY:-}
+      SKIP_PERMISSIONS: ${SKIP_PERMISSIONS:-}
+      TURNSTONE_AUTH_ENABLED: ${TURNSTONE_AUTH_ENABLED:-}
+      TURNSTONE_AUTH_TOKEN: ${TURNSTONE_AUTH_TOKEN:-}
+      TURNSTONE_JWT_SECRET: ${TURNSTONE_JWT_SECRET:-}
+      MODEL: ${MODEL:-}
+      MCP_CONFIG: ${MCP_CONFIG:-}
+      TURNSTONE_DB_BACKEND: ${DB_BACKEND:-postgresql}
+      TURNSTONE_DB_URL: ${DATABASE_URL:-postgresql://${POSTGRES_USER:-turnstone}:${POSTGRES_PASSWORD:?}@postgres:5432/turnstone}
+      TURNSTONE_NODE_ID: node-1
+    extra_hosts: ["host.docker.internal:host-gateway"]
+    networks: [turnstone-net]
+    depends_on:
+      redis: { condition: service_healthy }
+      postgres: { condition: service_healthy }
+      ddg-search: { condition: service_healthy, required: false }
+    healthcheck:
+      test: ["CMD", "python", "/usr/local/bin/healthcheck.py", "http://127.0.0.1:8080/health"]
+      interval: 10s
+      timeout: 5s
+      retries: 3
+      start_period: 15s
+    deploy:
+      resources:
+        limits: { memory: 384M, cpus: '0.5' }
+    restart: unless-stopped
+
+  server-2:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-2 }
+  server-3:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-3 }
+  server-4:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-4 }
+  server-5:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-5 }
+  server-6:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-6 }
+  server-7:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-7 }
+  server-8:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-8 }
+  server-9:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-9 }
+  server-10:
+    <<: *cluster-server
+    environment: { <<: *cluster-server-env, TURNSTONE_NODE_ID: node-10 }
+
+  # -- cluster bridges ------------------------------------------------
+
+  bridge-1: &cluster-bridge
+    build: { context: ., dockerfile: Dockerfile }
+    profiles: [cluster, ddgCluster]
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-1:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    environment: &cluster-bridge-env
+      REDIS_PASSWORD: ${REDIS_PASSWORD:-}
+      TURNSTONE_AUTH_TOKEN: ${TURNSTONE_AUTH_TOKEN:-}
+      TURNSTONE_JWT_SECRET: ${TURNSTONE_JWT_SECRET:-}
+    networks: [turnstone-net]
+    depends_on:
+      server-1: { condition: service_healthy }
+      redis: { condition: service_healthy }
+    deploy:
+      resources:
+        limits: { memory: 256M, cpus: '0.25' }
+    restart: unless-stopped
+
+  bridge-2:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-2:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-2: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-3:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-3:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-3: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-4:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-4:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-4: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-5:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-5:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-5: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-6:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-6:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-6: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-7:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-7:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-7: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-8:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-8:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-8: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-9:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-9:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-9: { condition: service_healthy }
+      redis: { condition: service_healthy }
+  bridge-10:
+    <<: *cluster-bridge
+    command:
+      - turnstone-bridge
+      - --server-url=http://server-10:8080
+      - --redis-host=redis
+      - --redis-port=6379
+      - --heartbeat-ttl=${HEARTBEAT_TTL:-60}
+      - --approval-timeout=${APPROVAL_TIMEOUT:-3600}
+    depends_on:
+      server-10: { condition: service_healthy }
+      redis: { condition: service_healthy }

demo.svg

@@ -1,221 +0,0 @@
-<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 860 520" font-family="ui-monospace,SFMono-Regular,Menlo,Monaco,Consolas,monospace" font-size="13">
-  <style>
-    @keyframes pulse-green  { 0%,100% { opacity:0.5 } 50% { opacity:1 } }
-    @keyframes pulse-yellow { 0%,100% { opacity:0.4 } 50% { opacity:1 } }
-    @keyframes pulse-blue   { 0%,100% { opacity:0.3 } 50% { opacity:1 } }
-    @keyframes fadein { from { opacity:0 } to { opacity:1 } }
-    .pg { animation: pulse-green 2s infinite }
-    .py { animation: pulse-yellow 1.8s infinite }
-    .pb { animation: pulse-blue 2.2s infinite }
-    .f1 { animation: fadein 0.4s 0.2s both }
-    .f2 { animation: fadein 0.4s 0.4s both }
-    .f3 { animation: fadein 0.4s 0.6s both }
-    .f4 { animation: fadein 0.4s 0.8s both }
-    .f5 { animation: fadein 0.4s 1.0s both }
-    .f6 { animation: fadein 0.4s 1.3s both }
-    .f7 { animation: fadein 0.4s 1.5s both }
-    .f8 { animation: fadein 0.4s 1.7s both }
-    .f9 { animation: fadein 0.4s 1.9s both }
-    .f10 { animation: fadein 0.4s 2.1s both }
-    .f11 { animation: fadein 0.4s 2.3s both }
-    .f12 { animation: fadein 0.4s 2.5s both }
-  </style>
-
-  <!-- Window chrome -->
-  <rect rx="10" width="860" height="520" fill="#1a1b26"/>
-  <rect width="860" height="36" rx="10" fill="#16161e"/>
-  <rect y="26" width="860" height="10" fill="#16161e"/>
-  <circle cx="20" cy="18" r="6" fill="#f7768e"/>
-  <circle cx="40" cy="18" r="6" fill="#e0af68"/>
-  <circle cx="60" cy="18" r="6" fill="#9ece6a"/>
-  <text x="430" y="22" text-anchor="middle" fill="#565f89" font-size="12">turnstone — console</text>
-
-  <!-- Header -->
-  <rect y="36" width="860" height="30" fill="#24283b"/>
-  <rect y="66" width="860" height="1" fill="#3b4261"/>
-  <text x="16" y="56" fill="#7aa2f7" font-size="14" font-weight="bold">turnstone console</text>
-  <text x="200" y="56" fill="#565f89" font-size="12">6 nodes · 10 workstreams</text>
-
-  <!-- ====== State cards ====== -->
-  <g transform="translate(16, 78)" class="f1" opacity="0">
-    <!-- RUN card -->
-    <rect x="0" y="0" width="156" height="64" rx="6" fill="#24283b" stroke="#3b4261"/>
-    <rect x="0" y="0" width="156" height="3" rx="6" fill="#9ece6a"/>
-    <text x="78" y="30" text-anchor="middle" fill="#a9b1d6" font-size="22" font-weight="bold">3</text>
-    <text x="78" y="50" text-anchor="middle" fill="#565f89" font-size="10">▸ RUN</text>
-
-    <!-- THINK card -->
-    <rect x="168" y="0" width="156" height="64" rx="6" fill="#24283b" stroke="#3b4261"/>
-    <rect x="168" y="0" width="156" height="3" rx="6" fill="#7aa2f7"/>
-    <text x="246" y="30" text-anchor="middle" fill="#a9b1d6" font-size="22" font-weight="bold">2</text>
-    <text x="246" y="50" text-anchor="middle" fill="#565f89" font-size="10">◌ THINK</text>
-
-    <!-- ATTN card -->
-    <rect x="336" y="0" width="156" height="64" rx="6" fill="#24283b" stroke="#3b4261"/>
-    <rect x="336" y="0" width="156" height="3" rx="6" fill="#e0af68"/>
-    <text x="414" y="30" text-anchor="middle" fill="#a9b1d6" font-size="22" font-weight="bold">1</text>
-    <text x="414" y="50" text-anchor="middle" fill="#565f89" font-size="10">◆ ATTN</text>
-
-    <!-- ERR card -->
-    <rect x="504" y="0" width="156" height="64" rx="6" fill="#24283b" stroke="#3b4261"/>
-    <rect x="504" y="0" width="156" height="3" rx="6" fill="#f7768e"/>
-    <text x="582" y="30" text-anchor="middle" fill="#a9b1d6" font-size="22" font-weight="bold">0</text>
-    <text x="582" y="50" text-anchor="middle" fill="#565f89" font-size="10">✖ ERR</text>
-
-    <!-- IDLE card -->
-    <rect x="672" y="0" width="156" height="64" rx="6" fill="#24283b" stroke="#3b4261"/>
-    <rect x="672" y="0" width="156" height="3" rx="6" fill="#565f89"/>
-    <text x="750" y="30" text-anchor="middle" fill="#a9b1d6" font-size="22" font-weight="bold">4</text>
-    <text x="750" y="50" text-anchor="middle" fill="#565f89" font-size="10">· IDLE</text>
-  </g>
-
-  <!-- Aggregate bar -->
-  <text x="16" y="160" fill="#565f89" font-size="11" class="f2" opacity="0">197k tokens · 42 tool calls</text>
-
-  <!-- ====== NODES section ====== -->
-  <text x="16" y="182" fill="#7aa2f7" font-size="12" font-weight="bold" class="f3" opacity="0">NODES</text>
-
-  <!-- Node column headers -->
-  <g transform="translate(0, 190)" class="f4" opacity="0">
-    <rect width="860" height="20" fill="#24283b"/>
-    <rect y="20" width="860" height="1" fill="#3b4261"/>
-    <text y="14" fill="#565f89" font-size="10" letter-spacing="0.5">
-      <tspan x="36">NODE</tspan>
-      <tspan x="560">WS</tspan>
-      <tspan x="610">RUN</tspan>
-      <tspan x="660">ATTN</tspan>
-      <tspan x="710">TOKENS</tspan>
-      <tspan x="790">LOAD</tspan>
-    </text>
-  </g>
-
-  <!-- Node rows -->
-  <g transform="translate(0, 214)">
-
-    <!-- Node 1: db-west-04 — 3 ws, 1 running, has-running bar -->
-    <g class="f5" opacity="0">
-      <rect y="0" width="860" height="38" fill="#1a1b26"/>
-      <rect y="0" width="3" height="38" fill="#9ece6a"/>
-      <circle cx="22" cy="19" r="4" fill="#9ece6a"/>
-      <text x="36" y="23" fill="#a9b1d6" font-size="12" font-weight="bold">db-west-04</text>
-      <text x="566" y="23" fill="#a9b1d6" font-size="11">3</text>
-      <text x="616" y="23" fill="#a9b1d6" font-size="11">1</text>
-      <text x="666" y="23" fill="#565f89" font-size="11">0</text>
-      <text x="710" y="23" fill="#565f89" font-size="11">57.6k</text>
-      <!-- Load bar: 3/10 = 30% -->
-      <rect x="770" y="15" width="60" height="6" rx="3" fill="#292e42"/>
-      <rect x="770" y="15" width="18" height="6" rx="3" fill="#9ece6a"/>
-      <text x="838" y="23" fill="#565f89" font-size="11">30%</text>
-    </g>
-
-    <!-- Node 2: api-east-01 — 3 ws, 1 attention, has-attention bar -->
-    <g class="f6" opacity="0">
-      <rect y="40" width="860" height="38" fill="#24283b"/>
-      <rect y="40" width="3" height="38" fill="#e0af68"/>
-      <circle cx="22" cy="59" r="4" fill="#9ece6a"/>
-      <text x="36" y="63" fill="#a9b1d6" font-size="12" font-weight="bold">api-east-01</text>
-      <text x="566" y="63" fill="#a9b1d6" font-size="11">3</text>
-      <text x="616" y="63" fill="#565f89" font-size="11">0</text>
-      <text x="666" y="63" fill="#a9b1d6" font-size="11">1</text>
-      <text x="710" y="63" fill="#565f89" font-size="11">109k</text>
-      <!-- Load bar: 3/10 = 30% -->
-      <rect x="770" y="55" width="60" height="6" rx="3" fill="#292e42"/>
-      <rect x="770" y="55" width="18" height="6" rx="3" fill="#9ece6a"/>
-      <text x="838" y="63" fill="#565f89" font-size="11">30%</text>
-    </g>
-
-    <!-- Node 3: sre-node-03 — 2 ws, 1 running, has-running bar -->
-    <g class="f7" opacity="0">
-      <rect y="80" width="860" height="38" fill="#1a1b26"/>
-      <rect y="80" width="3" height="38" fill="#9ece6a"/>
-      <circle cx="22" cy="99" r="4" fill="#9ece6a"/>
-      <text x="36" y="103" fill="#a9b1d6" font-size="12" font-weight="bold">sre-node-03</text>
-      <text x="566" y="103" fill="#a9b1d6" font-size="11">2</text>
-      <text x="616" y="103" fill="#a9b1d6" font-size="11">1</text>
-      <text x="666" y="103" fill="#565f89" font-size="11">0</text>
-      <text x="710" y="103" fill="#565f89" font-size="11">64.4k</text>
-      <!-- Load bar: 2/10 = 20% -->
-      <rect x="770" y="95" width="60" height="6" rx="3" fill="#292e42"/>
-      <rect x="770" y="95" width="12" height="6" rx="3" fill="#9ece6a"/>
-      <text x="838" y="103" fill="#565f89" font-size="11">20%</text>
-    </g>
-
-    <!-- Node 4: analytics-02 — 1 ws, thinking, has-thinking bar -->
-    <g class="f8" opacity="0">
-      <rect y="120" width="860" height="38" fill="#24283b"/>
-      <rect y="120" width="3" height="38" fill="#7aa2f7"/>
-      <circle cx="22" cy="139" r="4" fill="#9ece6a"/>
-      <text x="36" y="143" fill="#a9b1d6" font-size="12" font-weight="bold">analytics-02</text>
-      <text x="566" y="143" fill="#a9b1d6" font-size="11">1</text>
-      <text x="616" y="143" fill="#565f89" font-size="11">0</text>
-      <text x="666" y="143" fill="#565f89" font-size="11">0</text>
-      <text x="710" y="143" fill="#565f89" font-size="11">18.3k</text>
-      <!-- Load bar: 1/10 = 10% -->
-      <rect x="770" y="135" width="60" height="6" rx="3" fill="#292e42"/>
-      <rect x="770" y="135" width="6" height="6" rx="3" fill="#9ece6a"/>
-      <text x="838" y="143" fill="#565f89" font-size="11">10%</text>
-    </g>
-
-    <!-- Node 5: data-ops-05 — 1 ws, thinking, has-thinking bar -->
-    <g class="f9" opacity="0">
-      <rect y="160" width="860" height="38" fill="#1a1b26"/>
-      <rect y="160" width="3" height="38" fill="#7aa2f7"/>
-      <circle cx="22" cy="179" r="4" fill="#9ece6a"/>
-      <text x="36" y="183" fill="#a9b1d6" font-size="12" font-weight="bold">data-ops-05</text>
-      <text x="566" y="183" fill="#a9b1d6" font-size="11">1</text>
-      <text x="616" y="183" fill="#565f89" font-size="11">0</text>
-      <text x="666" y="183" fill="#565f89" font-size="11">0</text>
-      <text x="710" y="183" fill="#565f89" font-size="11">8.7k</text>
-      <!-- Load bar: 1/10 = 10% -->
-      <rect x="770" y="175" width="60" height="6" rx="3" fill="#292e42"/>
-      <rect x="770" y="175" width="6" height="6" rx="3" fill="#9ece6a"/>
-      <text x="838" y="183" fill="#565f89" font-size="11">10%</text>
-    </g>
-
-    <!-- Node 6: ml-gpu-07 — 0 ws, empty, no bar -->
-    <g class="f10" opacity="0">
-      <rect y="200" width="860" height="38" fill="#24283b"/>
-      <rect y="200" width="3" height="38" fill="transparent"/>
-      <circle cx="22" cy="219" r="4" fill="#9ece6a"/>
-      <text x="36" y="223" fill="#a9b1d6" font-size="12" font-weight="bold">ml-gpu-07</text>
-      <text x="566" y="223" fill="#565f89" font-size="11">0</text>
-      <text x="616" y="223" fill="#565f89" font-size="11">0</text>
-      <text x="666" y="223" fill="#565f89" font-size="11">0</text>
-      <text x="710" y="223" fill="#565f89" font-size="11">0</text>
-      <!-- Load bar: 0/10 = 0% (empty track) -->
-      <rect x="770" y="215" width="60" height="6" rx="3" fill="#292e42"/>
-      <text x="842" y="223" fill="#565f89" font-size="11">0%</text>
-    </g>
-
-  </g>
-
-  <!-- ====== Footer ====== -->
-  <g transform="translate(0, 468)" class="f12" opacity="0">
-    <rect width="860" height="1" fill="#3b4261"/>
-    <rect y="1" width="860" height="24" fill="#16161e"/>
-
-    <circle cx="20" cy="14" r="3" fill="#9ece6a"/>
-    <text x="28" y="18" fill="#565f89" font-size="10">db-west-04</text>
-
-    <circle cx="120" cy="14" r="3" fill="#9ece6a"/>
-    <text x="128" y="18" fill="#565f89" font-size="10">api-east-01</text>
-
-    <circle cx="225" cy="14" r="3" fill="#9ece6a"/>
-    <text x="233" y="18" fill="#565f89" font-size="10">sre-node-03</text>
-
-    <circle cx="335" cy="14" r="3" fill="#9ece6a"/>
-    <text x="343" y="18" fill="#565f89" font-size="10">analytics-02</text>
-
-    <circle cx="450" cy="14" r="3" fill="#9ece6a"/>
-    <text x="458" y="18" fill="#565f89" font-size="10">data-ops-05</text>
-
-    <circle cx="560" cy="14" r="3" fill="#9ece6a"/>
-    <text x="568" y="18" fill="#565f89" font-size="10">ml-gpu-07</text>
-
-    <text x="680" y="18" fill="#3b4261" font-size="10">258k tokens · 42 calls · 12m</text>
-  </g>
-
-  <!-- Bottom edge -->
-  <rect y="493" width="860" height="27" fill="#16161e"/>
-  <rect y="510" width="860" height="10" rx="10" fill="#16161e"/>
-</svg>

deploy/helm/turnstone/Chart.yaml

@@ -0,0 +1,16 @@
+apiVersion: v2
+name: turnstone
+description: Multi-node AI orchestration platform with tool use, agent routing, and cluster simulation
+type: application
+version: 0.1.0
+appVersion: "0.3.0"
+
+dependencies:
+  - name: postgresql
+    version: ~16.0
+    repository: https://charts.bitnami.com/bitnami
+    condition: postgresql.enabled
+  - name: redis
+    version: ~20.0
+    repository: https://charts.bitnami.com/bitnami
+    condition: redis.enabled

deploy/helm/turnstone/templates/NOTES.txt

@@ -0,0 +1,42 @@
+Turnstone {{ .Chart.AppVersion }} has been deployed.
+
+{{- if .Values.ingress.enabled }}
+
+Access the application via your ingress:
+{{- range .Values.ingress.hosts }}
+  http{{ if $.Values.ingress.tls }}s{{ end }}://{{ .host }}
+{{- end }}
+
+{{- else }}
+
+To access the Turnstone server, run:
+
+  kubectl port-forward svc/{{ include "turnstone.fullname" . }}-server {{ .Values.server.service.port }}:{{ .Values.server.service.port }}
+
+Then open: http://localhost:{{ .Values.server.service.port }}
+
+To access the Turnstone console (cluster dashboard), run:
+
+  kubectl port-forward svc/{{ include "turnstone.fullname" . }}-console {{ .Values.console.service.port }}:{{ .Values.console.service.port }}
+
+Then open: http://localhost:{{ .Values.console.service.port }}
+
+{{- end }}
+
+Components deployed:
+  - Server:  {{ include "turnstone.fullname" . }}-server  ({{ .Values.server.replicas }} replica(s))
+  - Bridge:  {{ include "turnstone.fullname" . }}-bridge  ({{ .Values.bridge.replicas }} replica(s))
+  - Console: {{ include "turnstone.fullname" . }}-console ({{ .Values.console.replicas }} replica(s))
+{{- if .Values.postgresql.enabled }}
+  - PostgreSQL (bitnami subchart)
+{{- end }}
+{{- if .Values.redis.enabled }}
+  - Redis (bitnami subchart)
+{{- end }}
+
+{{- if not .Values.llm.apiKey }}
+{{- if not .Values.llm.existingSecret }}
+
+WARNING: No LLM API key configured. Set llm.apiKey or llm.existingSecret in your values.
+{{- end }}
+{{- end }}

deploy/helm/turnstone/templates/_helpers.tpl

@@ -0,0 +1,163 @@
+{{/*
+Expand the name of the chart.
+*/}}
+{{- define "turnstone.name" -}}
+{{- default .Chart.Name .Values.nameOverride | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Create a default fully qualified app name.
+We truncate at 63 chars because some Kubernetes name fields are limited to this
+(by the DNS naming spec). If release name contains chart name it will be used
+as a full name.
+*/}}
+{{- define "turnstone.fullname" -}}
+{{- if .Values.fullnameOverride }}
+{{- .Values.fullnameOverride | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- $name := default .Chart.Name .Values.nameOverride }}
+{{- if contains $name .Release.Name }}
+{{- .Release.Name | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- printf "%s-%s" .Release.Name $name | trunc 63 | trimSuffix "-" }}
+{{- end }}
+{{- end }}
+{{- end }}
+
+{{/*
+Create chart name and version as used by the chart label.
+*/}}
+{{- define "turnstone.chart" -}}
+{{- printf "%s-%s" .Chart.Name .Chart.Version | replace "+" "_" | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Common labels.
+*/}}
+{{- define "turnstone.labels" -}}
+helm.sh/chart: {{ include "turnstone.chart" . }}
+{{ include "turnstone.selectorLabels" . }}
+{{- if .Chart.AppVersion }}
+app.kubernetes.io/version: {{ .Chart.AppVersion | quote }}
+{{- end }}
+app.kubernetes.io/managed-by: {{ .Release.Service }}
+{{- end }}
+
+{{/*
+Selector labels.
+*/}}
+{{- define "turnstone.selectorLabels" -}}
+app.kubernetes.io/name: {{ include "turnstone.name" . }}
+app.kubernetes.io/instance: {{ .Release.Name }}
+{{- end }}
+
+{{/*
+Create the name of the service account to use.
+*/}}
+{{- define "turnstone.serviceAccountName" -}}
+{{- if .Values.serviceAccount }}
+{{- if .Values.serviceAccount.name }}
+{{- .Values.serviceAccount.name }}
+{{- else }}
+{{- include "turnstone.fullname" . }}
+{{- end }}
+{{- else }}
+{{- include "turnstone.fullname" . }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the PostgreSQL host.
+*/}}
+{{- define "turnstone.postgresql.host" -}}
+{{- if .Values.postgresql.enabled }}
+{{- printf "%s-postgresql" .Release.Name }}
+{{- else }}
+{{- .Values.database.external.host }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the PostgreSQL port.
+*/}}
+{{- define "turnstone.postgresql.port" -}}
+{{- if .Values.postgresql.enabled }}
+{{- printf "5432" }}
+{{- else }}
+{{- .Values.database.external.port | toString }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the PostgreSQL database name.
+*/}}
+{{- define "turnstone.postgresql.database" -}}
+{{- if .Values.postgresql.enabled }}
+{{- .Values.postgresql.auth.database }}
+{{- else }}
+{{- .Values.database.external.database }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the PostgreSQL username.
+*/}}
+{{- define "turnstone.postgresql.username" -}}
+{{- if .Values.postgresql.enabled }}
+{{- .Values.postgresql.auth.username }}
+{{- else }}
+{{- .Values.database.external.username }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the Redis host.
+*/}}
+{{- define "turnstone.redis.host" -}}
+{{- if .Values.redis.enabled }}
+{{- printf "%s-redis-master" .Release.Name }}
+{{- else }}
+{{- .Values.redis.external.host }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the Redis port.
+*/}}
+{{- define "turnstone.redis.port" -}}
+{{- if .Values.redis.enabled }}
+{{- printf "6379" }}
+{{- else }}
+{{- .Values.redis.external.port | toString }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the secret name for LLM API keys.
+*/}}
+{{- define "turnstone.llm.secretName" -}}
+{{- if .Values.llm.existingSecret }}
+{{- .Values.llm.existingSecret }}
+{{- else }}
+{{- printf "%s-secrets" (include "turnstone.fullname" .) }}
+{{- end }}
+{{- end }}
+
+{{/*
+Determine the secret name for auth tokens.
+*/}}
+{{- define "turnstone.auth.secretName" -}}
+{{- if .Values.auth.existingSecret }}
+{{- .Values.auth.existingSecret }}
+{{- else }}
+{{- printf "%s-secrets" (include "turnstone.fullname" .) }}
+{{- end }}
+{{- end }}
+
+{{/*
+Container image reference.
+*/}}
+{{- define "turnstone.image" -}}
+{{- $tag := .Values.image.tag | default .Chart.AppVersion }}
+{{- printf "%s:%s" .Values.image.repository $tag }}
+{{- end }}

deploy/helm/turnstone/templates/configmap.yaml

@@ -0,0 +1,23 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: {{ include "turnstone.fullname" . }}-config
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+data:
+  TURNSTONE_DB_BACKEND: {{ .Values.database.backend | quote }}
+  TURNSTONE_DB_HOST: {{ include "turnstone.postgresql.host" . | quote }}
+  TURNSTONE_DB_PORT: {{ include "turnstone.postgresql.port" . | quote }}
+  TURNSTONE_DB_NAME: {{ include "turnstone.postgresql.database" . | quote }}
+  TURNSTONE_DB_USER: {{ include "turnstone.postgresql.username" . | quote }}
+  TURNSTONE_SERVER_HOST: "0.0.0.0"
+  TURNSTONE_SERVER_PORT: {{ .Values.server.service.port | quote }}
+  TURNSTONE_CONSOLE_HOST: "0.0.0.0"
+  TURNSTONE_CONSOLE_PORT: {{ .Values.console.service.port | quote }}
+  TURNSTONE_REDIS_HOST: {{ include "turnstone.redis.host" . | quote }}
+  TURNSTONE_REDIS_PORT: {{ include "turnstone.redis.port" . | quote }}
+  TURNSTONE_POLL_INTERVAL: "5"
+  {{- if .Values.llm.baseUrl }}
+  TURNSTONE_LLM_BASE_URL: {{ .Values.llm.baseUrl | quote }}
+  {{- end }}
+  TURNSTONE_LLM_PROVIDER: {{ .Values.llm.provider | quote }}

deploy/helm/turnstone/templates/deployment-bridge.yaml

@@ -0,0 +1,45 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: {{ include "turnstone.fullname" . }}-bridge
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: bridge
+spec:
+  replicas: {{ .Values.bridge.replicas }}
+  selector:
+    matchLabels:
+      {{- include "turnstone.selectorLabels" . | nindent 6 }}
+      app.kubernetes.io/component: bridge
+  template:
+    metadata:
+      labels:
+        {{- include "turnstone.selectorLabels" . | nindent 8 }}
+        app.kubernetes.io/component: bridge
+    spec:
+      serviceAccountName: {{ include "turnstone.serviceAccountName" . }}
+      containers:
+        - name: bridge
+          image: {{ include "turnstone.image" . }}
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          command:
+            - turnstone-bridge
+            - --server-url={{ printf "http://%s-server:%s" (include "turnstone.fullname" .) (.Values.server.service.port | toString) }}
+            - --redis-host={{ include "turnstone.redis.host" . }}
+            - --redis-port={{ include "turnstone.redis.port" . }}
+          envFrom:
+            - configMapRef:
+                name: {{ include "turnstone.fullname" . }}-config
+            - secretRef:
+                name: {{ include "turnstone.llm.secretName" . }}
+                optional: true
+          {{- if and .Values.auth.enabled .Values.auth.existingSecret }}
+          env:
+            - name: TURNSTONE_AUTH_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: {{ .Values.auth.existingSecret }}
+                  key: TURNSTONE_AUTH_TOKEN
+          {{- end }}
+          resources:
+            {{- toYaml .Values.bridge.resources | nindent 12 }}

deploy/helm/turnstone/templates/deployment-console.yaml

@@ -0,0 +1,62 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: {{ include "turnstone.fullname" . }}-console
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: console
+spec:
+  replicas: {{ .Values.console.replicas }}
+  selector:
+    matchLabels:
+      {{- include "turnstone.selectorLabels" . | nindent 6 }}
+      app.kubernetes.io/component: console
+  template:
+    metadata:
+      labels:
+        {{- include "turnstone.selectorLabels" . | nindent 8 }}
+        app.kubernetes.io/component: console
+    spec:
+      serviceAccountName: {{ include "turnstone.serviceAccountName" . }}
+      containers:
+        - name: console
+          image: {{ include "turnstone.image" . }}
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          command:
+            - turnstone-console
+            - --host=0.0.0.0
+            - --port={{ .Values.console.service.port }}
+            - --redis-host={{ include "turnstone.redis.host" . }}
+            - --redis-port={{ include "turnstone.redis.port" . }}
+          ports:
+            - name: http
+              containerPort: {{ .Values.console.service.port }}
+              protocol: TCP
+          envFrom:
+            - configMapRef:
+                name: {{ include "turnstone.fullname" . }}-config
+            - secretRef:
+                name: {{ include "turnstone.llm.secretName" . }}
+                optional: true
+          {{- if and .Values.auth.enabled .Values.auth.existingSecret }}
+          env:
+            - name: TURNSTONE_AUTH_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: {{ .Values.auth.existingSecret }}
+                  key: TURNSTONE_AUTH_TOKEN
+          {{- end }}
+          readinessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 10
+            periodSeconds: 10
+          livenessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 15
+            periodSeconds: 20
+          resources:
+            {{- toYaml .Values.console.resources | nindent 12 }}

deploy/helm/turnstone/templates/deployment-server.yaml

@@ -0,0 +1,64 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: {{ include "turnstone.fullname" . }}-server
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: server
+spec:
+  replicas: {{ .Values.server.replicas }}
+  selector:
+    matchLabels:
+      {{- include "turnstone.selectorLabels" . | nindent 6 }}
+      app.kubernetes.io/component: server
+  template:
+    metadata:
+      labels:
+        {{- include "turnstone.selectorLabels" . | nindent 8 }}
+        app.kubernetes.io/component: server
+    spec:
+      serviceAccountName: {{ include "turnstone.serviceAccountName" . }}
+      containers:
+        - name: server
+          image: {{ include "turnstone.image" . }}
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          command:
+            - turnstone-server
+            - --host
+            - "0.0.0.0"
+            - --port
+            - {{ .Values.server.service.port | quote }}
+          ports:
+            - name: http
+              containerPort: {{ .Values.server.service.port }}
+              protocol: TCP
+          envFrom:
+            - configMapRef:
+                name: {{ include "turnstone.fullname" . }}-config
+            - secretRef:
+                name: {{ include "turnstone.llm.secretName" . }}
+                optional: true
+          env:
+            - name: TURNSTONE_DB_URL
+              value: "postgresql+psycopg://$(TURNSTONE_DB_USER):$(POSTGRES_PASSWORD)@$(TURNSTONE_DB_HOST):$(TURNSTONE_DB_PORT)/$(TURNSTONE_DB_NAME)"
+          {{- if and .Values.auth.enabled .Values.auth.existingSecret }}
+            - name: TURNSTONE_AUTH_TOKEN
+              valueFrom:
+                secretKeyRef:
+                  name: {{ .Values.auth.existingSecret }}
+                  key: TURNSTONE_AUTH_TOKEN
+          {{- end }}
+          readinessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 10
+            periodSeconds: 10
+          livenessProbe:
+            httpGet:
+              path: /health
+              port: http
+            initialDelaySeconds: 15
+            periodSeconds: 20
+          resources:
+            {{- toYaml .Values.server.resources | nindent 12 }}

deploy/helm/turnstone/templates/ingress.yaml

@@ -0,0 +1,47 @@
+{{- if .Values.ingress.enabled -}}
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: {{ include "turnstone.fullname" . }}
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+  {{- with .Values.ingress.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+spec:
+  {{- if .Values.ingress.className }}
+  ingressClassName: {{ .Values.ingress.className }}
+  {{- end }}
+  {{- if .Values.ingress.tls }}
+  tls:
+    {{- range .Values.ingress.tls }}
+    - hosts:
+        {{- range .hosts }}
+        - {{ . | quote }}
+        {{- end }}
+      secretName: {{ .secretName }}
+    {{- end }}
+  {{- end }}
+  rules:
+    {{- range .Values.ingress.hosts }}
+    - host: {{ .host | quote }}
+      http:
+        paths:
+          {{- range .paths }}
+          - path: {{ .path }}
+            pathType: {{ .pathType | default "Prefix" }}
+            backend:
+              service:
+                {{- if eq (.service | default "server") "console" }}
+                name: {{ include "turnstone.fullname" $ }}-console
+                port:
+                  number: {{ $.Values.console.service.port }}
+                {{- else }}
+                name: {{ include "turnstone.fullname" $ }}-server
+                port:
+                  number: {{ $.Values.server.service.port }}
+                {{- end }}
+          {{- end }}
+    {{- end }}
+{{- end }}

deploy/helm/turnstone/templates/job-migrate.yaml

@@ -0,0 +1,38 @@
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: {{ include "turnstone.fullname" . }}-migrate
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: migrate
+  annotations:
+    "helm.sh/hook": pre-install,pre-upgrade
+    "helm.sh/hook-weight": "-1"
+    "helm.sh/hook-delete-policy": before-hook-creation,hook-succeeded
+spec:
+  backoffLimit: 3
+  template:
+    metadata:
+      labels:
+        {{- include "turnstone.selectorLabels" . | nindent 8 }}
+        app.kubernetes.io/component: migrate
+    spec:
+      serviceAccountName: {{ include "turnstone.serviceAccountName" . }}
+      restartPolicy: OnFailure
+      containers:
+        - name: migrate
+          image: {{ include "turnstone.image" . }}
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          command:
+            - python
+            - -m
+            - turnstone.core.storage._migrate
+          envFrom:
+            - configMapRef:
+                name: {{ include "turnstone.fullname" . }}-config
+            - secretRef:
+                name: {{ include "turnstone.llm.secretName" . }}
+                optional: true
+          env:
+            - name: TURNSTONE_DB_URL
+              value: "postgresql+psycopg://$(TURNSTONE_DB_USER):$(POSTGRES_PASSWORD)@$(TURNSTONE_DB_HOST):$(TURNSTONE_DB_PORT)/$(TURNSTONE_DB_NAME)"

deploy/helm/turnstone/templates/secret.yaml

@@ -0,0 +1,28 @@
+{{- if not .Values.llm.existingSecret }}
+apiVersion: v1
+kind: Secret
+metadata:
+  name: {{ include "turnstone.fullname" . }}-secrets
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+type: Opaque
+data:
+  {{- if .Values.llm.apiKey }}
+  OPENAI_API_KEY: {{ .Values.llm.apiKey | b64enc | quote }}
+  {{- end }}
+  {{- if and .Values.postgresql.enabled .Values.postgresql.auth.password }}
+  POSTGRES_PASSWORD: {{ .Values.postgresql.auth.password | b64enc | quote }}
+  {{- else if and (not .Values.postgresql.enabled) .Values.database.external.password }}
+  POSTGRES_PASSWORD: {{ .Values.database.external.password | b64enc | quote }}
+  {{- end }}
+  {{- if and .Values.auth.enabled .Values.auth.token (not .Values.auth.existingSecret) }}
+  TURNSTONE_AUTH_TOKEN: {{ .Values.auth.token | b64enc | quote }}
+  {{- end }}
+  {{- if and .Values.redis.enabled .Values.redis.auth }}
+  {{- if .Values.redis.auth.password }}
+  REDIS_PASSWORD: {{ .Values.redis.auth.password | b64enc | quote }}
+  {{- end }}
+  {{- else if and (not .Values.redis.enabled) .Values.redis.external.password }}
+  REDIS_PASSWORD: {{ .Values.redis.external.password | b64enc | quote }}
+  {{- end }}
+{{- end }}

deploy/helm/turnstone/templates/service-console.yaml

@@ -0,0 +1,17 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: {{ include "turnstone.fullname" . }}-console
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: console
+spec:
+  type: {{ .Values.console.service.type }}
+  ports:
+    - port: {{ .Values.console.service.port }}
+      targetPort: http
+      protocol: TCP
+      name: http
+  selector:
+    {{- include "turnstone.selectorLabels" . | nindent 4 }}
+    app.kubernetes.io/component: console

deploy/helm/turnstone/templates/service-server.yaml

@@ -0,0 +1,17 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: {{ include "turnstone.fullname" . }}-server
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}
+    app.kubernetes.io/component: server
+spec:
+  type: {{ .Values.server.service.type }}
+  ports:
+    - port: {{ .Values.server.service.port }}
+      targetPort: http
+      protocol: TCP
+      name: http
+  selector:
+    {{- include "turnstone.selectorLabels" . | nindent 4 }}
+    app.kubernetes.io/component: server

deploy/helm/turnstone/templates/serviceaccount.yaml

@@ -0,0 +1,6 @@
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ include "turnstone.serviceAccountName" . }}
+  labels:
+    {{- include "turnstone.labels" . | nindent 4 }}

deploy/helm/turnstone/values.yaml

@@ -0,0 +1,103 @@
+# -- Container image settings
+image:
+  repository: ghcr.io/turnstonelabs/turnstone
+  tag: ""
+  pullPolicy: IfNotPresent
+
+# -- Database configuration
+database:
+  # Backend type (postgresql)
+  backend: postgresql
+  # External database settings (used when postgresql.enabled is false)
+  external:
+    host: ""
+    port: 5432
+    database: turnstone
+    username: turnstone
+    existingSecret: ""
+    sslmode: prefer
+
+# -- Bitnami PostgreSQL subchart
+postgresql:
+  enabled: true
+  auth:
+    database: turnstone
+    username: turnstone
+
+# -- Redis configuration
+redis:
+  enabled: true
+  architecture: standalone
+  # External Redis settings (used when redis.enabled is false)
+  external:
+    host: ""
+    port: 6379
+    existingSecret: ""
+
+# -- Turnstone server (main API + web UI)
+server:
+  replicas: 1
+  resources:
+    requests:
+      cpu: 500m
+      memory: 512Mi
+    limits:
+      cpu: "2"
+      memory: 1Gi
+  service:
+    type: ClusterIP
+    port: 8080
+
+# -- Turnstone bridge (Redis MQ connector)
+bridge:
+  replicas: 1
+  resources:
+    requests:
+      cpu: 250m
+      memory: 256Mi
+    limits:
+      cpu: "1"
+      memory: 512Mi
+
+# -- Turnstone console (cluster dashboard)
+console:
+  replicas: 1
+  resources:
+    requests:
+      cpu: 250m
+      memory: 256Mi
+    limits:
+      cpu: "1"
+      memory: 512Mi
+  service:
+    type: ClusterIP
+    port: 8090
+
+# -- LLM provider configuration
+llm:
+  baseUrl: ""
+  provider: openai
+  apiKey: ""
+  existingSecret: ""
+
+# -- Authentication
+auth:
+  enabled: false
+  token: ""
+  existingSecret: ""
+
+# -- Ingress configuration
+ingress:
+  enabled: false
+  className: ""
+  annotations: {}
+  hosts: []
+  #  - host: turnstone.example.com
+  #    paths:
+  #      - path: /
+  #        pathType: Prefix
+  #        service: server
+  tls: []
+  #  - secretName: turnstone-tls
+  #    hosts:
+  #      - turnstone.example.com

deploy/terraform/examples/aws-ecs-basic/main.tf

@@ -0,0 +1,36 @@
+terraform {
+  required_version = ">= 1.5"
+
+  required_providers {
+    aws = {
+      source  = "hashicorp/aws"
+      version = ">= 5.0"
+    }
+  }
+}
+
+provider "aws" {
+  region = var.aws_region
+}
+
+module "turnstone" {
+  source = "../../modules/aws-ecs"
+
+  vpc_id             = var.vpc_id
+  private_subnet_ids = var.private_subnet_ids
+  public_subnet_ids  = var.public_subnet_ids
+
+  image_repository = var.image_repository
+  image_tag        = var.image_tag
+
+  llm_base_url   = var.llm_base_url
+  openai_api_key = var.openai_api_key
+
+  environment = var.environment
+  name_prefix = var.name_prefix
+  auth_token  = var.auth_token
+
+  tags = {
+    Example = "aws-ecs-basic"
+  }
+}

deploy/terraform/examples/aws-ecs-basic/outputs.tf

@@ -0,0 +1,29 @@
+output "alb_dns_name" {
+  description = "DNS name of the Application Load Balancer."
+  value       = module.turnstone.alb_dns_name
+}
+
+output "server_url" {
+  description = "HTTP URL for the Turnstone server."
+  value       = module.turnstone.server_url
+}
+
+output "console_url" {
+  description = "HTTP URL for the Turnstone console."
+  value       = module.turnstone.console_url
+}
+
+output "cluster_arn" {
+  description = "ARN of the ECS cluster."
+  value       = module.turnstone.cluster_arn
+}
+
+output "rds_endpoint" {
+  description = "RDS PostgreSQL endpoint."
+  value       = module.turnstone.rds_endpoint
+}
+
+output "redis_endpoint" {
+  description = "ElastiCache Redis endpoint."
+  value       = module.turnstone.redis_endpoint
+}

deploy/terraform/examples/aws-ecs-basic/terraform.tfvars.example

@@ -0,0 +1,22 @@
+# --- Required ---
+
+# VPC and subnet IDs from your existing AWS infrastructure.
+# The VPC must have DNS support and DNS hostnames enabled.
+vpc_id             = "vpc-0123456789abcdef0"
+private_subnet_ids = ["subnet-aaa111", "subnet-bbb222"]
+public_subnet_ids  = ["subnet-ccc333", "subnet-ddd444"]
+
+# LLM provider configuration.
+# For OpenAI: https://api.openai.com/v1
+# For a self-hosted vLLM instance: http://your-vllm-host:8000/v1
+llm_base_url   = "https://api.openai.com/v1"
+openai_api_key = "sk-..."
+
+# --- Optional ---
+
+# aws_region       = "us-east-1"
+# image_repository = "ghcr.io/turnstonelabs/turnstone"
+# image_tag        = "0.3.0"
+# environment      = "production"
+# name_prefix      = "turnstone"
+# auth_token       = "my-secret-token"

deploy/terraform/examples/aws-ecs-basic/variables.tf

@@ -0,0 +1,62 @@
+variable "aws_region" {
+  description = "AWS region to deploy into."
+  type        = string
+  default     = "us-east-1"
+}
+
+variable "vpc_id" {
+  description = "ID of the VPC where all resources will be created."
+  type        = string
+}
+
+variable "private_subnet_ids" {
+  description = "List of private subnet IDs for ECS tasks, RDS, and ElastiCache."
+  type        = list(string)
+}
+
+variable "public_subnet_ids" {
+  description = "List of public subnet IDs for the Application Load Balancer."
+  type        = list(string)
+}
+
+variable "image_repository" {
+  description = "Container image repository."
+  type        = string
+  default     = "ghcr.io/turnstonelabs/turnstone"
+}
+
+variable "image_tag" {
+  description = "Container image tag."
+  type        = string
+  default     = "latest"
+}
+
+variable "llm_base_url" {
+  description = "Base URL for the LLM provider API."
+  type        = string
+}
+
+variable "openai_api_key" {
+  description = "API key for the LLM provider."
+  type        = string
+  sensitive   = true
+}
+
+variable "environment" {
+  description = "Deployment environment name."
+  type        = string
+  default     = "production"
+}
+
+variable "name_prefix" {
+  description = "Prefix for all resource names."
+  type        = string
+  default     = "turnstone"
+}
+
+variable "auth_token" {
+  description = "Optional authentication token for the Turnstone API."
+  type        = string
+  sensitive   = true
+  default     = ""
+}

deploy/terraform/modules/aws-ecs/alb.tf

@@ -0,0 +1,150 @@
+# ---------- Application Load Balancer ----------
+#
+# HTTP listeners are provided as a starter baseline. For production, set
+# var.certificate_arn to an ACM certificate ARN to enable HTTPS listeners
+# that redirect HTTP traffic to TLS.
+
+resource "aws_lb" "this" {
+  name               = "${var.name_prefix}-${var.environment}"
+  internal           = false
+  load_balancer_type = "application"
+  security_groups    = [aws_security_group.alb.id]
+  subnets            = var.public_subnet_ids
+  tags               = local.common_tags
+}
+
+# ---------- Server Target Group + Listeners ----------
+
+resource "aws_lb_target_group" "server" {
+  name        = "${var.name_prefix}-server-${var.environment}"
+  port        = 8080
+  protocol    = "HTTP"
+  vpc_id      = var.vpc_id
+  target_type = "ip"
+  tags        = local.common_tags
+
+  health_check {
+    path                = "/health"
+    port                = "traffic-port"
+    protocol            = "HTTP"
+    healthy_threshold   = 2
+    unhealthy_threshold = 3
+    timeout             = 5
+    interval            = 30
+    matcher             = "200"
+  }
+}
+
+# HTTP listener: forwards directly when no certificate, redirects to HTTPS otherwise.
+resource "aws_lb_listener" "server" {
+  count             = var.certificate_arn == "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 80
+  protocol          = "HTTP"
+  tags              = local.common_tags
+
+  default_action {
+    type             = "forward"
+    target_group_arn = aws_lb_target_group.server.arn
+  }
+}
+
+resource "aws_lb_listener" "server_http_redirect" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 80
+  protocol          = "HTTP"
+  tags              = local.common_tags
+
+  default_action {
+    type = "redirect"
+    redirect {
+      port        = "443"
+      protocol    = "HTTPS"
+      status_code = "HTTP_301"
+    }
+  }
+}
+
+resource "aws_lb_listener" "server_https" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 443
+  protocol          = "HTTPS"
+  ssl_policy        = "ELBSecurityPolicy-TLS13-1-2-2021-06"
+  certificate_arn   = var.certificate_arn
+  tags              = local.common_tags
+
+  default_action {
+    type             = "forward"
+    target_group_arn = aws_lb_target_group.server.arn
+  }
+}
+
+# ---------- Console Target Group + Listeners ----------
+
+resource "aws_lb_target_group" "console" {
+  name        = "${var.name_prefix}-console-${var.environment}"
+  port        = 8090
+  protocol    = "HTTP"
+  vpc_id      = var.vpc_id
+  target_type = "ip"
+  tags        = local.common_tags
+
+  health_check {
+    path                = "/health"
+    port                = "traffic-port"
+    protocol            = "HTTP"
+    healthy_threshold   = 2
+    unhealthy_threshold = 3
+    timeout             = 5
+    interval            = 30
+    matcher             = "200"
+  }
+}
+
+# HTTP listener: forwards directly when no certificate, redirects to HTTPS otherwise.
+resource "aws_lb_listener" "console" {
+  count             = var.certificate_arn == "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 8090
+  protocol          = "HTTP"
+  tags              = local.common_tags
+
+  default_action {
+    type             = "forward"
+    target_group_arn = aws_lb_target_group.console.arn
+  }
+}
+
+resource "aws_lb_listener" "console_http_redirect" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 8090
+  protocol          = "HTTP"
+  tags              = local.common_tags
+
+  default_action {
+    type = "redirect"
+    redirect {
+      port        = "8443"
+      protocol    = "HTTPS"
+      status_code = "HTTP_301"
+    }
+  }
+}
+
+resource "aws_lb_listener" "console_https" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  load_balancer_arn = aws_lb.this.arn
+  port              = 8443
+  protocol          = "HTTPS"
+  ssl_policy        = "ELBSecurityPolicy-TLS13-1-2-2021-06"
+  certificate_arn   = var.certificate_arn
+  tags              = local.common_tags
+
+  default_action {
+    type             = "forward"
+    target_group_arn = aws_lb_target_group.console.arn
+  }
+}

deploy/terraform/modules/aws-ecs/elasticache.tf

@@ -0,0 +1,30 @@
+# ---------- ElastiCache Subnet Group ----------
+
+resource "aws_elasticache_subnet_group" "this" {
+  name       = "${var.name_prefix}-${var.environment}"
+  subnet_ids = var.private_subnet_ids
+  tags       = local.common_tags
+}
+
+# ---------- ElastiCache Redis Replication Group ----------
+
+resource "aws_elasticache_replication_group" "this" {
+  replication_group_id = "${var.name_prefix}-${var.environment}"
+  description          = "Turnstone Redis for MQ and session state"
+
+  engine               = "redis"
+  engine_version       = "7.1"
+  node_type            = var.redis_node_type
+  num_cache_clusters   = 1
+  port                 = 6379
+
+  subnet_group_name  = aws_elasticache_subnet_group.this.name
+  security_group_ids = [aws_security_group.redis.id]
+
+  at_rest_encryption_enabled = true
+  transit_encryption_enabled = true
+
+  automatic_failover_enabled = false
+
+  tags = local.common_tags
+}

deploy/terraform/modules/aws-ecs/iam.tf

@@ -0,0 +1,70 @@
+# ---------- ECS Task Execution Role ----------
+# Used by the ECS agent to pull images and retrieve secrets.
+
+resource "aws_iam_role" "ecs_execution" {
+  name = "${var.name_prefix}-ecs-execution-${var.environment}"
+  tags = local.common_tags
+
+  assume_role_policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [
+      {
+        Effect = "Allow"
+        Principal = {
+          Service = "ecs-tasks.amazonaws.com"
+        }
+        Action = "sts:AssumeRole"
+      },
+    ]
+  })
+}
+
+resource "aws_iam_role_policy_attachment" "ecs_execution_base" {
+  role       = aws_iam_role.ecs_execution.name
+  policy_arn = "arn:aws:iam::aws:policy/service-role/AmazonECSTaskExecutionRolePolicy"
+}
+
+resource "aws_iam_role_policy" "ecs_execution_secrets" {
+  name = "${var.name_prefix}-secrets-read-${var.environment}"
+  role = aws_iam_role.ecs_execution.id
+
+  policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [
+      {
+        Effect = "Allow"
+        Action = [
+          "secretsmanager:GetSecretValue",
+        ]
+        Resource = concat(
+          [
+            aws_secretsmanager_secret.openai_api_key.arn,
+            aws_secretsmanager_secret.db_password.arn,
+          ],
+          var.auth_token != "" ? [aws_secretsmanager_secret.auth_token[0].arn] : [],
+        )
+      },
+    ]
+  })
+}
+
+# ---------- ECS Task Role ----------
+# Assumed by the running container. Minimal permissions; extend as needed.
+
+resource "aws_iam_role" "ecs_task" {
+  name = "${var.name_prefix}-ecs-task-${var.environment}"
+  tags = local.common_tags
+
+  assume_role_policy = jsonencode({
+    Version = "2012-10-17"
+    Statement = [
+      {
+        Effect = "Allow"
+        Principal = {
+          Service = "ecs-tasks.amazonaws.com"
+        }
+        Action = "sts:AssumeRole"
+      },
+    ]
+  })
+}

deploy/terraform/modules/aws-ecs/main.tf

@@ -0,0 +1,313 @@
+terraform {
+  required_version = ">= 1.5"
+
+  required_providers {
+    aws = {
+      source  = "hashicorp/aws"
+      version = ">= 5.0"
+    }
+    random = {
+      source  = "hashicorp/random"
+      version = ">= 3.5"
+    }
+  }
+}
+
+locals {
+  full_image = "${var.image_repository}:${var.image_tag}"
+
+  common_tags = merge(var.tags, {
+    Project     = "turnstone"
+    Environment = var.environment
+    ManagedBy   = "terraform"
+  })
+
+  # Shared environment variables injected into every container.
+  common_env = [
+    { name = "TURNSTONE_ENV", value = var.environment },
+    { name = "TURNSTONE_DB_BACKEND", value = "postgresql" },
+    { name = "TURNSTONE_LLM_BASE_URL", value = var.llm_base_url },
+    { name = "TURNSTONE_REDIS_URL", value = "redis://${aws_elasticache_replication_group.this.primary_endpoint_address}:6379/0" },
+  ]
+
+  # Secrets pulled from Secrets Manager at container start.
+  common_secrets = [
+    {
+      name      = "OPENAI_API_KEY"
+      valueFrom = aws_secretsmanager_secret_version.openai_api_key.arn
+    },
+    {
+      name      = "TURNSTONE_DB_URL"
+      valueFrom = aws_secretsmanager_secret_version.db_url.arn
+    },
+  ]
+
+  auth_env = var.auth_token != "" ? [
+    { name = "TURNSTONE_AUTH_ENABLED", value = "true" },
+  ] : []
+
+  auth_secrets = var.auth_token != "" ? [
+    {
+      name      = "TURNSTONE_AUTH_TOKEN"
+      valueFrom = aws_secretsmanager_secret_version.auth_token[0].arn
+    },
+  ] : []
+}
+
+# ---------- Secrets Manager ----------
+
+resource "aws_secretsmanager_secret" "openai_api_key" {
+  name = "${var.name_prefix}-${var.environment}-openai-api-key"
+  tags = local.common_tags
+}
+
+resource "aws_secretsmanager_secret_version" "openai_api_key" {
+  secret_id     = aws_secretsmanager_secret.openai_api_key.id
+  secret_string = var.openai_api_key
+}
+
+resource "aws_secretsmanager_secret" "auth_token" {
+  count = var.auth_token != "" ? 1 : 0
+  name  = "${var.name_prefix}-${var.environment}-auth-token"
+  tags  = local.common_tags
+}
+
+resource "aws_secretsmanager_secret_version" "auth_token" {
+  count         = var.auth_token != "" ? 1 : 0
+  secret_id     = aws_secretsmanager_secret.auth_token[0].id
+  secret_string = var.auth_token
+}
+
+resource "aws_secretsmanager_secret" "db_password" {
+  name = "${var.name_prefix}-${var.environment}-db-password"
+  tags = local.common_tags
+}
+
+resource "aws_secretsmanager_secret_version" "db_password" {
+  secret_id     = aws_secretsmanager_secret.db_password.id
+  secret_string = random_password.db.result
+}
+
+resource "aws_secretsmanager_secret" "db_url" {
+  name = "${var.name_prefix}-${var.environment}-db-url"
+  tags = local.common_tags
+}
+
+resource "aws_secretsmanager_secret_version" "db_url" {
+  secret_id     = aws_secretsmanager_secret.db_url.id
+  secret_string = "postgresql+psycopg://${aws_db_instance.this.username}:${random_password.db.result}@${aws_db_instance.this.endpoint}/turnstone"
+}
+
+# ---------- ECS Cluster ----------
+
+resource "aws_ecs_cluster" "this" {
+  name = "${var.name_prefix}-${var.environment}"
+  tags = local.common_tags
+
+  setting {
+    name  = "containerInsights"
+    value = "enabled"
+  }
+}
+
+# ---------- CloudWatch Log Group ----------
+
+resource "aws_cloudwatch_log_group" "this" {
+  name              = "/ecs/${var.name_prefix}-${var.environment}"
+  retention_in_days = 30
+  tags              = local.common_tags
+}
+
+# ---------- Server Task Definition + Service ----------
+
+resource "aws_ecs_task_definition" "server" {
+  family                   = "${var.name_prefix}-server"
+  requires_compatibilities = ["FARGATE"]
+  network_mode             = "awsvpc"
+  cpu                      = var.server_cpu
+  memory                   = var.server_memory
+  execution_role_arn       = aws_iam_role.ecs_execution.arn
+  task_role_arn            = aws_iam_role.ecs_task.arn
+  tags                     = local.common_tags
+
+  container_definitions = jsonencode([
+    {
+      name      = "server"
+      image     = local.full_image
+      essential = true
+      command   = ["turnstone-server", "--host", "0.0.0.0", "--port", "8080"]
+
+      portMappings = [
+        { containerPort = 8080, protocol = "tcp" },
+      ]
+
+      environment = concat(local.common_env, local.auth_env)
+      secrets     = concat(local.common_secrets, local.auth_secrets)
+
+      logConfiguration = {
+        logDriver = "awslogs"
+        options = {
+          "awslogs-group"         = aws_cloudwatch_log_group.this.name
+          "awslogs-region"        = data.aws_region.current.name
+          "awslogs-stream-prefix" = "server"
+        }
+      }
+
+      healthCheck = {
+        command     = ["CMD-SHELL", "curl -f http://localhost:8080/health || exit 1"]
+        interval    = 30
+        timeout     = 5
+        retries     = 3
+        startPeriod = 10
+      }
+    },
+  ])
+}
+
+resource "aws_ecs_service" "server" {
+  name            = "${var.name_prefix}-server"
+  cluster         = aws_ecs_cluster.this.id
+  task_definition = aws_ecs_task_definition.server.arn
+  desired_count   = 1
+  launch_type     = "FARGATE"
+  tags            = local.common_tags
+
+  network_configuration {
+    subnets          = var.private_subnet_ids
+    security_groups  = [aws_security_group.ecs_tasks.id]
+    assign_public_ip = false
+  }
+
+  load_balancer {
+    target_group_arn = aws_lb_target_group.server.arn
+    container_name   = "server"
+    container_port   = 8080
+  }
+
+  depends_on = [aws_lb_target_group.server]
+}
+
+# ---------- Bridge Task Definition + Service ----------
+
+resource "aws_ecs_task_definition" "bridge" {
+  family                   = "${var.name_prefix}-bridge"
+  requires_compatibilities = ["FARGATE"]
+  network_mode             = "awsvpc"
+  cpu                      = var.bridge_cpu
+  memory                   = var.bridge_memory
+  execution_role_arn       = aws_iam_role.ecs_execution.arn
+  task_role_arn            = aws_iam_role.ecs_task.arn
+  tags                     = local.common_tags
+
+  container_definitions = jsonencode([
+    {
+      name      = "bridge"
+      image     = local.full_image
+      essential = true
+      command   = ["turnstone-bridge"]
+
+      environment = concat(local.common_env, local.auth_env)
+      secrets     = concat(local.common_secrets, local.auth_secrets)
+
+      logConfiguration = {
+        logDriver = "awslogs"
+        options = {
+          "awslogs-group"         = aws_cloudwatch_log_group.this.name
+          "awslogs-region"        = data.aws_region.current.name
+          "awslogs-stream-prefix" = "bridge"
+        }
+      }
+    },
+  ])
+}
+
+resource "aws_ecs_service" "bridge" {
+  name            = "${var.name_prefix}-bridge"
+  cluster         = aws_ecs_cluster.this.id
+  task_definition = aws_ecs_task_definition.bridge.arn
+  desired_count   = 1
+  launch_type     = "FARGATE"
+  tags            = local.common_tags
+
+  network_configuration {
+    subnets          = var.private_subnet_ids
+    security_groups  = [aws_security_group.ecs_tasks.id]
+    assign_public_ip = false
+  }
+
+  depends_on = [aws_ecs_service.server]
+}
+
+# ---------- Console Task Definition + Service ----------
+
+resource "aws_ecs_task_definition" "console" {
+  family                   = "${var.name_prefix}-console"
+  requires_compatibilities = ["FARGATE"]
+  network_mode             = "awsvpc"
+  cpu                      = var.console_cpu
+  memory                   = var.console_memory
+  execution_role_arn       = aws_iam_role.ecs_execution.arn
+  task_role_arn            = aws_iam_role.ecs_task.arn
+  tags                     = local.common_tags
+
+  container_definitions = jsonencode([
+    {
+      name      = "console"
+      image     = local.full_image
+      essential = true
+      command   = ["turnstone-console", "--host", "0.0.0.0", "--port", "8090"]
+
+      portMappings = [
+        { containerPort = 8090, protocol = "tcp" },
+      ]
+
+      environment = concat(local.common_env, local.auth_env)
+      secrets     = concat(local.common_secrets, local.auth_secrets)
+
+      logConfiguration = {
+        logDriver = "awslogs"
+        options = {
+          "awslogs-group"         = aws_cloudwatch_log_group.this.name
+          "awslogs-region"        = data.aws_region.current.name
+          "awslogs-stream-prefix" = "console"
+        }
+      }
+
+      healthCheck = {
+        command     = ["CMD-SHELL", "curl -f http://localhost:8090/health || exit 1"]
+        interval    = 30
+        timeout     = 5
+        retries     = 3
+        startPeriod = 10
+      }
+    },
+  ])
+}
+
+resource "aws_ecs_service" "console" {
+  name            = "${var.name_prefix}-console"
+  cluster         = aws_ecs_cluster.this.id
+  task_definition = aws_ecs_task_definition.console.arn
+  desired_count   = 1
+  launch_type     = "FARGATE"
+  tags            = local.common_tags
+
+  network_configuration {
+    subnets          = var.private_subnet_ids
+    security_groups  = [aws_security_group.ecs_tasks.id]
+    assign_public_ip = false
+  }
+
+  load_balancer {
+    target_group_arn = aws_lb_target_group.console.arn
+    container_name   = "console"
+    container_port   = 8090
+  }
+
+  depends_on = [aws_lb_target_group.console]
+}
+
+# ---------- Data Sources ----------
+
+data "aws_region" "current" {}
+data "aws_caller_identity" "current" {}

deploy/terraform/modules/aws-ecs/outputs.tf

@@ -0,0 +1,29 @@
+output "alb_dns_name" {
+  description = "DNS name of the Application Load Balancer."
+  value       = aws_lb.this.dns_name
+}
+
+output "server_url" {
+  description = "HTTP URL for the Turnstone server API and web UI."
+  value       = "http://${aws_lb.this.dns_name}"
+}
+
+output "console_url" {
+  description = "HTTP URL for the Turnstone console dashboard."
+  value       = "http://${aws_lb.this.dns_name}:8090"
+}
+
+output "cluster_arn" {
+  description = "ARN of the ECS cluster."
+  value       = aws_ecs_cluster.this.arn
+}
+
+output "rds_endpoint" {
+  description = "Endpoint of the RDS PostgreSQL instance (host:port)."
+  value       = aws_db_instance.this.endpoint
+}
+
+output "redis_endpoint" {
+  description = "Primary endpoint of the ElastiCache Redis replication group."
+  value       = aws_elasticache_replication_group.this.primary_endpoint_address
+}

deploy/terraform/modules/aws-ecs/rds.tf

@@ -0,0 +1,42 @@
+# ---------- Random Password ----------
+
+resource "random_password" "db" {
+  length  = 32
+  special = false
+}
+
+# ---------- DB Subnet Group ----------
+
+resource "aws_db_subnet_group" "this" {
+  name       = "${var.name_prefix}-${var.environment}"
+  subnet_ids = var.private_subnet_ids
+  tags       = local.common_tags
+}
+
+# ---------- RDS PostgreSQL ----------
+
+resource "aws_db_instance" "this" {
+  identifier = "${var.name_prefix}-${var.environment}"
+
+  engine               = "postgres"
+  engine_version       = "17"
+  instance_class       = var.db_instance_class
+  allocated_storage    = 20
+  storage_type         = "gp3"
+  storage_encrypted    = true
+  deletion_protection  = true
+  skip_final_snapshot  = false
+  final_snapshot_identifier = "${var.name_prefix}-${var.environment}-final"
+
+  db_name  = "turnstone"
+  username = "turnstone"
+  password = random_password.db.result
+
+  db_subnet_group_name   = aws_db_subnet_group.this.name
+  vpc_security_group_ids = [aws_security_group.rds.id]
+
+  backup_retention_period = 7
+  multi_az                = false
+
+  tags = local.common_tags
+}

deploy/terraform/modules/aws-ecs/security.tf

@@ -0,0 +1,133 @@
+# ---------- ALB Security Group ----------
+
+resource "aws_security_group" "alb" {
+  name        = "${var.name_prefix}-alb-${var.environment}"
+  description = "Allow inbound HTTP to ALB for server and console"
+  vpc_id      = var.vpc_id
+  tags        = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "alb_http" {
+  security_group_id = aws_security_group.alb.id
+  description       = "HTTP traffic to server"
+  from_port         = 80
+  to_port           = 80
+  ip_protocol       = "tcp"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "alb_https" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  security_group_id = aws_security_group.alb.id
+  description       = "HTTPS traffic to server"
+  from_port         = 443
+  to_port           = 443
+  ip_protocol       = "tcp"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "alb_console" {
+  security_group_id = aws_security_group.alb.id
+  description       = "HTTP traffic to console"
+  from_port         = 8090
+  to_port           = 8090
+  ip_protocol       = "tcp"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "alb_console_https" {
+  count             = var.certificate_arn != "" ? 1 : 0
+  security_group_id = aws_security_group.alb.id
+  description       = "HTTPS traffic to console"
+  from_port         = 8443
+  to_port           = 8443
+  ip_protocol       = "tcp"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+resource "aws_vpc_security_group_egress_rule" "alb_all" {
+  security_group_id = aws_security_group.alb.id
+  description       = "Allow all outbound"
+  ip_protocol       = "-1"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+# ---------- ECS Tasks Security Group ----------
+
+resource "aws_security_group" "ecs_tasks" {
+  name        = "${var.name_prefix}-ecs-tasks-${var.environment}"
+  description = "Allow traffic from ALB to ECS tasks and outbound internet"
+  vpc_id      = var.vpc_id
+  tags        = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "ecs_from_alb_server" {
+  security_group_id            = aws_security_group.ecs_tasks.id
+  description                  = "Server port from ALB"
+  from_port                    = 8080
+  to_port                      = 8080
+  ip_protocol                  = "tcp"
+  referenced_security_group_id = aws_security_group.alb.id
+  tags                         = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "ecs_from_alb_console" {
+  security_group_id            = aws_security_group.ecs_tasks.id
+  description                  = "Console port from ALB"
+  from_port                    = 8090
+  to_port                      = 8090
+  ip_protocol                  = "tcp"
+  referenced_security_group_id = aws_security_group.alb.id
+  tags                         = local.common_tags
+}
+
+resource "aws_vpc_security_group_egress_rule" "ecs_all" {
+  security_group_id = aws_security_group.ecs_tasks.id
+  description       = "Allow all outbound (LLM APIs, ECR, Secrets Manager, etc.)"
+  ip_protocol       = "-1"
+  cidr_ipv4         = "0.0.0.0/0"
+  tags              = local.common_tags
+}
+
+# ---------- RDS Security Group ----------
+
+resource "aws_security_group" "rds" {
+  name        = "${var.name_prefix}-rds-${var.environment}"
+  description = "Allow PostgreSQL access from ECS tasks"
+  vpc_id      = var.vpc_id
+  tags        = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "rds_from_ecs" {
+  security_group_id            = aws_security_group.rds.id
+  description                  = "PostgreSQL from ECS tasks"
+  from_port                    = 5432
+  to_port                      = 5432
+  ip_protocol                  = "tcp"
+  referenced_security_group_id = aws_security_group.ecs_tasks.id
+  tags                         = local.common_tags
+}
+
+# ---------- Redis Security Group ----------
+
+resource "aws_security_group" "redis" {
+  name        = "${var.name_prefix}-redis-${var.environment}"
+  description = "Allow Redis access from ECS tasks"
+  vpc_id      = var.vpc_id
+  tags        = local.common_tags
+}
+
+resource "aws_vpc_security_group_ingress_rule" "redis_from_ecs" {
+  security_group_id            = aws_security_group.redis.id
+  description                  = "Redis from ECS tasks"
+  from_port                    = 6379
+  to_port                      = 6379
+  ip_protocol                  = "tcp"
+  referenced_security_group_id = aws_security_group.ecs_tasks.id
+  tags                         = local.common_tags
+}

deploy/terraform/modules/aws-ecs/variables.tf

@@ -0,0 +1,130 @@
+# --- Networking ---
+
+variable "vpc_id" {
+  description = "ID of the VPC where all resources will be created."
+  type        = string
+}
+
+variable "private_subnet_ids" {
+  description = "List of private subnet IDs for ECS tasks, RDS, and ElastiCache."
+  type        = list(string)
+}
+
+variable "public_subnet_ids" {
+  description = "List of public subnet IDs for the Application Load Balancer."
+  type        = list(string)
+}
+
+# --- Container Image ---
+
+variable "image_repository" {
+  description = "Container image repository."
+  type        = string
+  default     = "ghcr.io/turnstonelabs/turnstone"
+}
+
+variable "image_tag" {
+  description = "Container image tag."
+  type        = string
+  default     = "latest"
+}
+
+# --- LLM Provider ---
+
+variable "llm_base_url" {
+  description = "Base URL for the LLM provider API (e.g. https://api.openai.com/v1)."
+  type        = string
+}
+
+variable "openai_api_key" {
+  description = "API key for the LLM provider. Stored in AWS Secrets Manager."
+  type        = string
+  sensitive   = true
+}
+
+# --- RDS ---
+
+variable "db_instance_class" {
+  description = "RDS instance class for PostgreSQL."
+  type        = string
+  default     = "db.t4g.micro"
+}
+
+# --- ElastiCache ---
+
+variable "redis_node_type" {
+  description = "ElastiCache node type for Redis."
+  type        = string
+  default     = "cache.t4g.micro"
+}
+
+# --- ECS Task Sizing ---
+
+variable "server_cpu" {
+  description = "CPU units for the server task (1 vCPU = 1024)."
+  type        = number
+  default     = 512
+}
+
+variable "server_memory" {
+  description = "Memory (MiB) for the server task."
+  type        = number
+  default     = 1024
+}
+
+variable "bridge_cpu" {
+  description = "CPU units for the bridge task."
+  type        = number
+  default     = 256
+}
+
+variable "bridge_memory" {
+  description = "Memory (MiB) for the bridge task."
+  type        = number
+  default     = 512
+}
+
+variable "console_cpu" {
+  description = "CPU units for the console task."
+  type        = number
+  default     = 256
+}
+
+variable "console_memory" {
+  description = "Memory (MiB) for the console task."
+  type        = number
+  default     = 512
+}
+
+# --- General ---
+
+variable "environment" {
+  description = "Deployment environment name (e.g. production, staging)."
+  type        = string
+  default     = "production"
+}
+
+variable "name_prefix" {
+  description = "Prefix for all resource names."
+  type        = string
+  default     = "turnstone"
+}
+
+variable "auth_token" {
+  description = "Optional authentication token for the Turnstone API. Empty string disables auth."
+  type        = string
+  sensitive   = true
+  default     = ""
+}
+
+variable "certificate_arn" {
+  description = "ACM certificate ARN for HTTPS listeners. Leave empty for HTTP-only (not recommended for production)."
+  type        = string
+  default     = ""
+}
+
+variable "tags" {
+  description = "Additional tags to apply to all resources."
+  type        = map(string)
+  default     = {}
+}

docker/entrypoint.sh

@@ -0,0 +1,5 @@
+#!/bin/sh
+# Run database migrations before starting the service
+python -m turnstone.core.storage._migrate || true
+# Execute the actual command
+exec "$@"

docker/mcp-ddg.json

@@ -0,0 +1,7 @@
+{
+  "mcpServers": {
+    "ddg": {
+      "url": "http://ddg-search:3000/mcp"
+    }
+  }
+}

docs/channels.md

@@ -0,0 +1,346 @@
+# Channel Integrations
+
+The `turnstone-channel` gateway connects external messaging platforms to
+turnstone workstreams via Redis MQ. Each platform adapter translates
+platform-native events (messages, button clicks, slash commands) into
+turnstone MQ messages, and renders workstream output back into the
+platform's UI.
+
+Discord ships as the first adapter. The adapter protocol is designed for
+future Slack and Teams integrations.
+
+---
+
+## Architecture
+
+```
+Discord Gateway
+      |
+      v
+turnstone-channel  (Discord adapter)
+      |
+      v
+  Redis MQ
+      |
+      v
+turnstone-bridge  ──>  turnstone-server
+```
+
+Key components:
+
+- **ChannelAdapter protocol** (`turnstone/channels/_protocol.py`) — generic
+  interface for any messaging platform. Defines `start()`, `stop()`,
+  `send()`, `edit_message()`, `send_approval_request()`,
+  `send_plan_review()`, and `create_thread()`.
+- **ChannelRouter** (`turnstone/channels/_routing.py`) — maps
+  channel/thread IDs to turnstone workstream IDs. Handles workstream
+  creation via MQ, stale route detection, and user identity resolution.
+- **AsyncRedisBroker** (`turnstone/mq/async_broker.py`) — async Redis
+  client compatible with discord.py's event loop. Used by the router for
+  pub/sub and queue operations.
+- **channel_users table** — maps `(channel_type, channel_user_id)` to a
+  turnstone `user_id`. Messages from unlinked users are silently dropped.
+- **channel_routes table** — persistent channel-to-workstream mappings.
+  Survives bot restarts. Stale routes (evicted workstreams) are detected
+  and refreshed on the next message.
+
+---
+
+## Discord Setup
+
+### 1. Create a Discord Application
+
+1. Go to https://discord.com/developers/applications
+2. Click **New Application** and give it a name
+3. Navigate to the **Bot** tab and click **Reset Token** to generate a
+   bot token. Copy it immediately — it is shown only once.
+4. On the same **Bot** tab, scroll down to **Privileged Gateway Intents**
+   and enable **MESSAGE CONTENT INTENT**
+5. Navigate to **OAuth2 > URL Generator**
+6. Under **Scopes**, check `bot` and `applications.commands`
+7. Under **Bot Permissions**, check:
+   - View Channels
+   - Send Messages
+   - Send Messages in Threads
+   - Create Public Threads
+   - Read Message History
+   - Add Reactions
+   - Embed Links
+8. Copy the generated URL, open it in a browser, and add the bot to your
+   Discord server
+
+### 2. Configure Turnstone
+
+**Environment variables** (recommended for Docker):
+
+```bash
+TURNSTONE_DISCORD_TOKEN=your-bot-token-here
+TURNSTONE_DISCORD_GUILD=123456789  # optional, restrict to one guild
+```
+
+**CLI flags** (bare-metal):
+
+```bash
+turnstone-channel \
+  --discord-token "your-bot-token" \
+  --discord-guild 123456789 \
+  --redis-host localhost \
+  --redis-port 6379
+```
+
+**Docker Compose** (production profile):
+
+```bash
+# In .env file:
+TURNSTONE_DISCORD_TOKEN=your-bot-token
+TURNSTONE_DISCORD_GUILD=123456789
+```
+
+Then start the stack:
+
+```bash
+docker compose --profile production up
+```
+
+The `channel` service starts automatically when
+`TURNSTONE_DISCORD_TOKEN` is set.
+
+### 3. Link User Accounts
+
+Discord users must link their account to a turnstone user before they can
+interact with the bot. Unlinked users' messages are silently ignored.
+
+1. The user must have a turnstone API token — created via the admin panel
+   or `turnstone-admin create-token`
+2. In Discord, the user runs `/link`. A modal appears prompting for the
+   API token (the token is never visible in Discord audit logs because it
+   is submitted via modal, not as a slash command argument).
+3. The token is validated against the database. If valid, a
+   `channel_users` mapping is created.
+4. The user can now @mention the bot or use slash commands.
+
+An admin can also force-link or unlink users via the console admin panel
+(Admin > Channels tab).
+
+---
+
+## Usage
+
+### Conversations
+
+- **@mention** the bot in any allowed channel to start a new conversation.
+  The bot creates a Discord thread from the message and a turnstone
+  workstream behind it.
+- All subsequent messages in the thread are routed to the same workstream.
+- The bot streams responses via message edits, updated approximately every
+  1.5 seconds.
+- If the workstream is evicted for capacity, the next message in the
+  thread auto-creates a new workstream and atomically resumes the
+  previous workstream via the `resume_ws` field on
+  `CreateWorkstreamMessage`. The server resumes the workstream during
+  creation (same HTTP request), and the bridge emits a
+  `WorkstreamResumedEvent` back to the channel. The thread receives a
+  *"Resumed: {name} ({count} messages restored)"* confirmation.
+
+### Slash Commands
+
+| Command | Description |
+|---------|-------------|
+| `/link` | Link Discord account to turnstone (opens modal for API token) |
+| `/unlink` | Unlink Discord account |
+| `/ask <message>` | Create a new thread and workstream with an initial message |
+| `/status` | Show workstream info for the current thread (ephemeral) |
+| `/close` | Close the workstream, delete the route, and archive the thread |
+
+### Tool Approvals
+
+When manual approval is enabled (the default), tool calls are displayed as
+an orange embed with:
+
+- Tool name and argument preview
+- **Approve** (green), **Reject** (red), **Always Approve** (gray) buttons
+- Only linked users can interact with approval buttons
+- The approval decision is forwarded through MQ to the bridge, which
+  relays it to the server
+
+Buttons use static `custom_id` values so they survive bot restarts.
+Correlation data (`ws_id`, `correlation_id`) is stored in the embed footer.
+
+**Auto-approval:** When `auto_approve` is true (via `--auto-approve`), or when
+all tools in the request match the `auto_approve_tools` list in the adapter
+config, the bot auto-responds with approval and posts a
+"*Tool auto-approved.*" notice to the thread instead of showing buttons. The
+`auto_approve_tools` list is set via the `ChannelConfig.auto_approve_tools`
+field (useful for allowing specific tools like `bash` or `read_file` while
+still requiring manual approval for others).
+
+### Plan Reviews
+
+Plan review requests are displayed as a blue embed with:
+
+- **Approve Plan** (green) button — approves the plan with empty feedback
+- **Request Changes** (gray) button — opens a modal for feedback text
+  (up to 2000 characters)
+- Feedback is forwarded through MQ as a `PlanFeedbackMessage`
+
+---
+
+## Configuration Reference
+
+| CLI Flag | Env Var | Default | Description |
+|----------|---------|---------|-------------|
+| `--discord-token` | `TURNSTONE_DISCORD_TOKEN` | — | Bot token (required to enable Discord) |
+| `--discord-guild` | — | `0` (all guilds) | Restrict to a single Discord guild |
+| `--discord-channels` | — | empty (all) | Comma-separated channel IDs to allow |
+| `--redis-host` | `REDIS_HOST` | `localhost` | Redis host |
+| `--redis-port` | — | `6379` | Redis port |
+| `--redis-password` | `REDIS_PASSWORD` | — | Redis password |
+| `--redis-db` | — | `0` | Redis DB number |
+| `--model` | — | server default | Default model for new workstreams |
+| `--auto-approve` | — | `false` | Auto-approve ALL tool calls (skips approval buttons entirely) |
+| `--http-host` | — | `127.0.0.1` | HTTP server bind address for notify endpoint |
+| `--http-port` | `TURNSTONE_CHANNEL_PORT` | `8091` | HTTP server port |
+| `--auth-token` | `TURNSTONE_CHANNEL_AUTH_TOKEN` | — | Static auth token for `/v1/api/notify` (alternative to JWT) |
+| `--log-level` | `TURNSTONE_LOG_LEVEL` | `INFO` | Log level |
+| `--log-format` | `TURNSTONE_LOG_FORMAT` | `auto` | Log format (`auto`/`json`/`text`) |
+
+---
+
+## User Identity
+
+- The `channel_users` table maps `(channel_type, channel_user_id)` to a
+  turnstone `user_id`
+- Self-service linking via the `/link` slash command (modal input, not
+  visible in Discord audit logs)
+- Admin can force-link or unlink via the console admin panel (Admin >
+  Channels tab). Unlinking uses a styled confirmation modal.
+- Unlinked users' messages are silently dropped
+- A user can be linked across multiple platforms (e.g. Discord + Slack)
+
+See [Security: Database Schema](security.md#database-schema) for the
+`channel_users` table definition.
+
+---
+
+## Workstream Lifecycle
+
+1. **Creation** — @mention or `/ask` creates a Discord thread and a
+   turnstone workstream. The `ChannelRouter` persists the mapping in the
+   `channel_routes` table.
+2. **Active** — messages are routed bidirectionally. The bot streams
+   responses via message edits (updated every ~1.5 seconds).
+3. **Eviction** — the server evicts an idle workstream for capacity. The
+   route is preserved and the thread stays open.
+4. **Reactivation** — the next message in the thread detects the stale
+   route (no MQ owner) and creates a new workstream with the old `ws_id`
+   as `resume_ws` on the `CreateWorkstreamMessage`. The server resumes
+   the workstream during creation (no separate command or reverse lookup
+   needed). The bridge emits a `WorkstreamResumedEvent` to the channel, and
+   the thread displays *"Resumed: {name} ({count} messages restored)"*.
+   If the old workstream was pruned, a fresh one starts with no error.
+5. **Close** — `/close` command closes the workstream via MQ, deletes the
+   route, unsubscribes from events, and archives the Discord thread.
+
+---
+
+## Notifications
+
+> See also: [Notification Flow diagram](diagrams/png/17-notify-flow.png)
+
+The `notify` tool allows the LLM to proactively send notifications to
+users or channels on external platforms. This is useful for alerting
+people about task completion, errors, or important updates without
+waiting for them to check in.
+
+### Targeting
+
+Two modes:
+
+- **Username** — provide a turnstone `username`. The gateway resolves
+  it via the `channel_users` table and sends to all linked channels
+  (e.g. Discord + future Slack).
+- **Direct** — provide `channel_type` + `channel_id` to target a
+  specific platform channel or user DM.
+
+### Delivery Flow
+
+Notifications bypass MQ for lower latency. The server calls the channel
+gateway directly over HTTP:
+
+1. The LLM calls the `notify` tool with a message and target
+2. `_exec_notify()` queries the `services` table for healthy channel
+   gateways (heartbeat within the last 120 seconds)
+3. The server mints a service JWT (`aud: turnstone-channel`) via
+   `ServiceTokenManager` and POSTs to the first healthy gateway
+4. The gateway validates the JWT, resolves the target, and calls
+   `adapter.send()` on the appropriate platform adapter
+5. On failure, the server tries the next gateway. If all fail, it
+   retries up to 2 more times (delays: 1s, 3s), re-querying the
+   service registry on each attempt
+
+### Service Registry
+
+The channel gateway registers itself in the `services` database table
+on startup and sends a heartbeat every 30 seconds. On shutdown it
+deregisters. Services are considered stale after 120 seconds (4 missed
+heartbeats) and are excluded from `list_services()` queries.
+
+The `services` table schema:
+
+| Column | Description |
+|--------|-------------|
+| `service_type` | Service category (e.g. `"channel"`) |
+| `service_id` | Unique instance ID (`channel-<hostname>-<random>`) |
+| `url` | HTTP base URL for the service |
+| `last_heartbeat` | ISO 8601 timestamp of last heartbeat |
+| `created` | ISO 8601 timestamp of initial registration |
+
+### Security
+
+- **Authentication** — the gateway's `POST /v1/api/notify` endpoint
+  requires authentication. Configure either `TURNSTONE_JWT_SECRET`
+  (the server mints JWTs with `aud: turnstone-channel` automatically)
+  or a static token via `--auth-token`. If neither is set, the
+  gateway fails closed and rejects all requests with 401. Server JWTs
+  (`aud: turnstone-server`) are rejected.
+- **Rate limit** — maximum 5 notifications per turn. The counter only
+  increments on successful delivery, so failures don't consume the
+  budget.
+- **SSRF protection** — only `http://` and `https://` service URLs
+  are allowed. Other schemes are silently skipped.
+- **Mention sanitization** — `discord.utils.escape_mentions()` is
+  applied before sending, preventing `@everyone` / `@here` abuse.
+- **Error redaction** — generic error messages are returned to the
+  LLM. Internal details (service IDs, URLs, exception messages) are
+  logged server-side only.
+
+---
+
+## Adding New Adapters
+
+The `ChannelAdapter` protocol defines the interface any platform adapter
+must implement:
+
+```python
+class ChannelAdapter(Protocol):
+    channel_type: str
+
+    async def start(self) -> None: ...
+    async def stop(self) -> None: ...
+    async def send(self, channel_id: str, content: str) -> str: ...
+    async def edit_message(self, channel_id: str, message_id: str, content: str) -> None: ...
+    async def send_approval_request(self, channel_id: str, ws_id: str, correlation_id: str, items: list[dict]) -> None: ...
+    async def send_plan_review(self, channel_id: str, ws_id: str, correlation_id: str, content: str) -> None: ...
+    async def create_thread(self, parent_channel_id: str, name: str, message_id: str = "") -> str: ...
+```
+
+To add a new platform:
+
+1. Create `turnstone/channels/<platform>/` package
+2. Implement the `ChannelAdapter` protocol
+3. Add a `--<platform>-token` flag and detection logic in
+   `turnstone/channels/cli.py`
+4. Add the optional dependency in `pyproject.toml` (e.g.
+   `turnstone[slack]`)
+
+See `turnstone/channels/discord/` as a reference implementation.

docs/console.md

@@ -1,28 +1,40 @@
 # Cluster Dashboard (turnstone-console)
 
-`turnstone-console` is a standalone monitoring service that provides cluster-wide visibility across all turnstone nodes. It connects to the shared Redis broker, discovers nodes via heartbeat keys, polls each node's HTTP API for workstream data, and subscribes to a cluster event channel for real-time state changes.
+`turnstone-console` is a cluster management service that provides cluster-wide visibility and control across all turnstone nodes. It connects to the shared Redis broker, discovers nodes via heartbeat keys, polls each node's HTTP API for workstream data, and subscribes to a cluster event channel for real-time state changes.
 
-The console is read-only — it observes but does not own workstreams or drive LLM sessions.
+The console also supports **workstream creation** (dispatched via MQ to target nodes) and a **reverse proxy** that serves each node's server UI through the console port — so users only need network access to the console, not to individual server nodes.
 
 ## Architecture
 
 > See also: [Console Data Flow diagram](diagrams/png/11-console-data-flow.png)
 
 ```
-turnstone-server ──→ turnstone-bridge ──→ Redis ──→ turnstone-console ──→ Browser
-     (per node)         (per node)       (shared)      (one instance)
+                        ┌── Redis ←── turnstone-bridge ←── turnstone-server
+                        │    (MQ)        (per node)           (per node)
+turnstone-console ──────┤
+   (one instance)       │
+                        └── turnstone-server (direct HTTP proxy)
+        │
+        ▼
+     Browser
 ```
 
-Each bridge publishes state changes to `{prefix}:events:cluster` on Redis pub/sub. The console subscribes once to that channel for real-time updates and periodically polls each node's `GET /api/dashboard` for full workstream snapshots.
+Data flows in two directions:
+
+- **Inbound (monitoring):** Bridges publish state changes to `{prefix}:events:cluster` on Redis pub/sub. The console subscribes for real-time updates and periodically polls each node's `GET /v1/api/dashboard` for full workstream snapshots.
+- **Outbound (control):** The console pushes `CreateWorkstreamMessage` to Redis inbound queues targeting specific nodes. Bridges pick up these messages and create workstreams on their local servers.
+- **Proxy (pass-through):** The console reverse-proxies each node's server UI at `/node/{node_id}/`, forwarding HTTP and SSE traffic so the browser never contacts server nodes directly.
 
 ### Data Sources
 
-| Source | Method | Frequency | Data |
+| Source | Method | Direction | Data |
 |--------|--------|-----------|------|
-| Redis heartbeats | `SCAN turnstone:node:*` | Every 15s | Node discovery (node_id, server_url, started) |
-| Redis pub/sub | `SUBSCRIBE turnstone:events:cluster` | Real-time | State changes, creates, closes, renames |
-| Node HTTP API | `GET {server_url}/api/dashboard` | Every 10s | Full workstream list with tokens, context, activity |
-| Node HTTP API | `GET {server_url}/health` | Every 10s | Node health status |
+| Redis heartbeats | `SCAN turnstone:node:*` | Read | Node discovery (node_id, server_url, started) |
+| Redis pub/sub | `SUBSCRIBE turnstone:events:cluster` | Read | State changes, creates, closes, renames |
+| Node HTTP API | `GET {server_url}/v1/api/dashboard` | Read | Full workstream list with tokens, context, activity |
+| Node HTTP API | `GET {server_url}/health` | Read | Node health status |
+| Redis inbound queue | `RPUSH turnstone:inbound:{node_id}` | Write | Workstream creation commands |
+| Node HTTP API | `GET/POST {server_url}/*` | Proxy | Server UI, API requests, SSE streams |
 
 ### Redis Key: Cluster Event Channel
 
@@ -47,7 +59,9 @@ The collector (`turnstone/console/collector.py`) maintains an in-memory snapshot
 
 2. **Node discovery** — scans heartbeat keys every 15 seconds via `broker.list_nodes()`. Adds newly discovered nodes, removes expired ones, emits `node_joined` / `node_lost` events to SSE listeners.
 
-3. **Poll loop** — fetches `GET /api/dashboard` and `GET /health` from each known node every 10 seconds. Uses `ThreadPoolExecutor(max_workers=50)` for parallelism. Each poll replaces the node's workstream list with the authoritative server data.
+3. **Poll loop** — fetches `GET /v1/api/dashboard` and `GET /health` from each known node every 10 seconds. Uses `ThreadPoolExecutor(max_workers=50)` for parallelism. Each poll replaces the node's workstream list with the authoritative server data.
+
+A `get_snapshot()` method builds the full cluster state under a single lock acquisition — overview aggregates and per-node workstream lists in one atomic read. This is served both as a REST endpoint and as the initial SSE event on client connect.
 
 ### Thread Safety
 
@@ -64,7 +78,7 @@ All reads and writes to the node/workstream map are protected by a single `threa
 
 ## HTTP API
 
-### `GET /api/cluster/overview`
+### `GET /v1/api/cluster/overview`
 
 Cluster-wide state counts and aggregate metrics.
 
@@ -73,11 +87,15 @@ Cluster-wide state counts and aggregate metrics.
   "nodes": 847,
   "workstreams": 4219,
   "states": {"running": 1847, "thinking": 312, "attention": 89, "idle": 1940, "error": 31},
-  "aggregate": {"total_tokens": 12400000, "total_tool_calls": 34200}
+  "aggregate": {"total_tokens": 12400000, "total_tool_calls": 34200},
+  "version_drift": true,
+  "versions": ["0.3.0", "0.3.1"]
 }
 ```
 
-### `GET /api/cluster/nodes?sort=activity&limit=100&offset=0`
+`version_drift` is `true` when nodes report different versions. `versions` lists all unique version strings sorted alphabetically.
+
+### `GET /v1/api/cluster/nodes?sort=activity&limit=100&offset=0`
 
 Paginated node list. Sort options: `activity` (default, by running+attention count), `tokens`, `name`.
 
@@ -91,14 +109,15 @@ Paginated node list. Sort options: `activity` (default, by running+attention cou
       "total_tokens": 48200,
       "started": 1709294400.0,
       "reachable": true,
-      "health": {}
+      "health": {"status": "ok", "version": "0.3.0"},
+      "version": "0.3.0"
     }
   ],
   "total": 847
 }
 ```
 
-### `GET /api/cluster/workstreams?state=running&node=db-west-04&search=perf&page=1&per_page=50`
+### `GET /v1/api/cluster/workstreams?state=running&node=db-west-04&search=perf&page=1&per_page=50`
 
 Filtered, paginated workstream list. All query parameters are optional. `per_page` is capped at 200.
 
@@ -115,7 +134,7 @@ Filtered, paginated workstream list. All query parameters are optional. `per_pag
 }
 ```
 
-### `GET /api/cluster/node/{node_id}`
+### `GET /v1/api/cluster/node/{node_id}`
 
 Single node detail with all its workstreams.
 
@@ -129,9 +148,75 @@ Single node detail with all its workstreams.
 }
 ```
 
-### `GET /api/cluster/events`
+### `GET /v1/api/cluster/snapshot`
 
-Server-Sent Events stream for real-time cluster updates.
+Full cluster state in a single response — all nodes with their workstreams plus overview aggregates. Built under a single lock for internal consistency. Used by the browser on initial load and SSE reconnect.
+
+```json
+{
+  "nodes": [
+    {
+      "node_id": "db-west-04",
+      "server_url": "http://10.0.3.4:8080",
+      "max_ws": 10,
+      "reachable": true,
+      "version": "0.3.0",
+      "health": {"status": "ok", "version": "0.3.0"},
+      "aggregate": {"total_tokens": 48200, "total_tool_calls": 156},
+      "workstreams": [
+        {"id": "a1b2c3d4", "name": "perf-db-west", "state": "running", ...}
+      ]
+    }
+  ],
+  "overview": {
+    "nodes": 847,
+    "workstreams": 4219,
+    "states": {"running": 1847, "thinking": 312, "attention": 89, "idle": 1940, "error": 31},
+    "aggregate": {"total_tokens": 12400000, "total_tool_calls": 34200},
+    "version_drift": false,
+    "versions": ["0.3.0"]
+  },
+  "timestamp": 1709294400.0
+}
+```
+
+### `POST /v1/api/cluster/workstreams/new`
+
+Create a new workstream on a target node. Dispatches a `CreateWorkstreamMessage` through the Redis MQ pipeline — the bridge on the target node picks it up and creates the workstream on the server. Requires `write` scope.
+
+Request:
+
+```json
+{
+  "node_id": "db-west-04",
+  "name": "perf-analysis",
+  "model": "gpt-5"
+}
+```
+
+All fields are optional:
+- `node_id` — targeting mode:
+  - **omitted or `"auto"`** — console picks the reachable node with the most available capacity (max_ws - ws_total) and pushes to its directed queue.
+  - **`"pool"`** — pushes to the shared inbound queue; the next available bridge picks it up (true general-pool dispatch).
+  - **specific node ID** — pushes to that node's directed queue.
+- `name` — workstream display name. Auto-generated if omitted.
+- `model` — model alias from the target node's registry. Uses the node's default model if omitted.
+
+Response:
+
+```json
+{
+  "status": "ok",
+  "correlation_id": "a1b2c3d4e5f6",
+  "target_node": "db-west-04"
+}
+```
+
+Creation is asynchronous — the response confirms the MQ message was dispatched. A `ws_created` event on the cluster SSE stream confirms the workstream was actually created.
+
+### `GET /v1/api/cluster/events`
+
+Server-Sent Events stream for real-time cluster updates. The first event is always a `snapshot` containing the full cluster state (same shape as `GET /v1/api/cluster/snapshot` with an added `type: "snapshot"` field), followed by incremental events:
 
 ```
 data: {"type":"cluster_state","ws_id":"a1b2","node_id":"db-west-04","state":"running"}
@@ -146,32 +231,390 @@ Keepalive comments (`: keepalive\n\n`) are sent every 5 seconds. Clients should
 ### `GET /health`
 
 ```json
-{"status": "ok", "service": "turnstone-console", "nodes": 847, "workstreams": 4219}
+{
+  "status": "ok",
+  "service": "turnstone-console",
+  "nodes": 847,
+  "workstreams": 4219,
+  "version_drift": false,
+  "versions": ["0.3.0"]
+}
 ```
 
+### Admin API
+
+User and token management endpoints. All admin endpoints require `approve` scope, except for the setup endpoint which is public.
+
+#### `POST /v1/api/auth/setup`
+
+Creates the first admin user when no users exist. Public endpoint (no auth required). Returns a JWT and sets a session cookie. Returns `409` if users already exist. See [Security: First-time setup](security.md#first-time-setup) for full details.
+
+#### `POST /v1/api/admin/users`
+
+Create a new user.
+
+```json
+{
+  "username": "alice",
+  "password": "s3cret",
+  "scopes": ["read", "write"]
+}
+```
+
+#### `GET /v1/api/admin/users`
+
+List all users.
+
+```json
+{
+  "users": [
+    {"user_id": "u_abc123", "username": "alice", "scopes": ["read", "write"], "created": "2026-03-01T12:00:00Z"}
+  ]
+}
+```
+
+#### `DELETE /v1/api/admin/users/{user_id}`
+
+Delete a user and revoke all their tokens.
+
+#### `POST /v1/api/admin/users/{user_id}/tokens`
+
+Create an API token for the given user. Returns a `ts_`-prefixed token string that can be used for Bearer auth or passed to `client.login(token="ts_xxx")`.
+
+```json
+{
+  "name": "CI pipeline",
+  "scopes": ["read", "write"]
+}
+```
+
+#### `GET /v1/api/admin/users/{user_id}/tokens`
+
+List active tokens for a user (token strings are not returned, only metadata).
+
+#### `DELETE /v1/api/admin/tokens/{token_id}`
+
+Revoke a specific API token.
+
+### Channel links
+
+| Method | Path | Description |
+|--------|------|-------------|
+| GET | `/v1/api/admin/users/{user_id}/channels` | List channel links for a user |
+| POST | `/v1/api/admin/users/{user_id}/channels` | Link a channel account (channel_type, channel_user_id) |
+| DELETE | `/v1/api/admin/channels/{channel_type}/{channel_user_id}` | Unlink a channel account |
+
+These endpoints manage the `channel_users` table mappings that connect external platform identities (e.g. Discord user IDs) to turnstone users. See [Channel Integrations](channels.md) for details on the linking flow.
+
+### Workstream Templates
+
+| Method | Path | Description |
+|--------|------|-------------|
+| GET | `/v1/api/admin/ws-templates` | List all workstream templates |
+| POST | `/v1/api/admin/ws-templates` | Create a workstream template |
+| GET | `/v1/api/admin/ws-templates/{id}` | Get a single workstream template |
+| PUT | `/v1/api/admin/ws-templates/{id}` | Update (auto-versions, audit logged) |
+| DELETE | `/v1/api/admin/ws-templates/{id}` | Delete + cascade versions (audit logged) |
+| GET | `/v1/api/admin/ws-templates/{id}/versions` | Version history |
+| GET | `/v1/api/ws-templates` | Enabled templates summary (name, description, model) — requires write scope, not admin |
+
+#### `GET /v1/api/auth/status`
+
+Public endpoint for login UI state detection. Returns auth configuration, not
+current-user identity.
+
+```json
+{
+  "auth_enabled": true,
+  "has_users": true,
+  "setup_required": false
+}
+```
+
+### Auth Scopes
+
+The auth system uses three scopes instead of the earlier read/full role model:
+
+| Scope | Grants |
+|-------|--------|
+| `read` | Read-only access: dashboards, workstream lists, SSE streams, health |
+| `write` | Send messages, create/close workstreams, approve tool calls |
+| `approve` | Admin operations: manage users and API tokens |
+
+Scopes are cumulative — a user with `approve` scope can also perform `write` and `read` operations.
+
+---
+
+## Reverse Proxy
+
+The console reverse-proxies each node's server UI at `/node/{node_id}/`. This allows users to interact with any node's workstreams through the console port alone — individual server ports do not need to be exposed to the office network.
+
+### Proxy Routes
+
+| Route | Behavior |
+|-------|----------|
+| `GET /node/{node_id}/` | Fetches the server's `index.html`, rewrites static and shared asset paths, injects a console-return banner and an inline JS proxy shim |
+| `GET /node/{node_id}/static/{path}` | Proxies page-specific static files |
+| `GET /node/{node_id}/shared/{path}` | Proxies shared static files (`base.css`, `auth.js`, etc.) |
+| `GET /node/{node_id}/v1/api/{path}` | Proxies GET API requests; detects SSE endpoints and streams them |
+| `POST /node/{node_id}/v1/api/{path}` | Proxies POST API requests with body forwarding |
+| `GET /node/{node_id}/{path}` | Proxies non-API endpoints (health, metrics) |
+
+### URL Rewriting
+
+The server UI uses root-relative URLs (`/v1/api/send`, `/static/app.js`, `/shared/base.css`, etc.). Since `<base>` tags cannot rewrite root-relative URLs, the console uses a JS shim approach:
+
+1. **HTML rewriting** — when serving `index.html`, replaces `href=` and `src=` references to both `/static/` and `/shared/` with the proxy prefix (`/node/{node_id}/static/` and `/node/{node_id}/shared/` respectively).
+
+2. **Inline JS shim** — injects an inline `<script>` block into the proxied HTML (after the console-return banner, before any external scripts) that overrides `window.fetch()` and `window.EventSource()` to prepend the proxy prefix to any root-relative URL. Running the shim inline ensures it executes before any external scripts load, so all API calls and SSE connections are intercepted transparently.
+
+3. **Console-return banner** — injects a thin inline-styled `<div>` after `<body>` with a "← Console" link and the node ID, providing navigation back to the dashboard.
+
+### SSE Proxy
+
+SSE streams (`/v1/api/events`, `/v1/api/events/global`) are proxied as raw byte passthrough — the console opens an `httpx.AsyncClient.stream()` to the upstream server (with `read=None` and `pool=None` timeouts since SSE connections are long-lived) and relays every byte via `StreamingResponse`. This preserves server-side ping comments, event framing, and keepalives verbatim without parsing or re-encoding.
+
+### Authentication
+
+The proxy forwards the user's JWT to upstream server nodes — it extracts the token from the incoming request's cookie (or `Authorization` header) and adds it as a `Bearer` header on the proxied request.  Since all services share the same `TURNSTONE_JWT_SECRET`, the user's JWT is valid on every node without re-authentication.  The console's own auth middleware also checks proxy routes — `POST` requests to proxy write endpoints (`/v1/api/send`, `/v1/api/approve`, etc.) require `write` scope, preventing read-only tokens from escalating via proxy.  The static `--auth-token` / `proxy_auth_token` is used as a fallback when no user JWT is present.
+
 ---
 
 ## Browser Dashboard
 
-The web UI has three views, toggled client-side:
+The web UI has five views, toggled client-side:
 
 ### 1. Cluster Overview (landing)
 
 - **State cards** — 5 clickable cards (running, thinking, attention, idle, error) with count and colored top border. Clicking filters to that state.
 - **Aggregate bar** — total tokens and tool calls across the cluster.
-- **Node table** — columns: NODE, WS, RUN, ATTN, TOKENS, HEALTH. Sorted by activity. Clickable rows drill down to node detail.
+- **Node table** — columns: NODE, WS, RUN, ATTN, TOKENS, VER, LOAD. Sorted by activity. Clickable rows drill down to node detail. Version column shows per-node version; hidden on mobile.
+- **Version drift indicator** — when nodes report different versions, the status bar shows a yellow "DRIFT" warning with a tooltip listing all versions. Node groups show "mixed" with a yellow badge when their members disagree.
+- **"+ new" button** — opens the workstream creation modal (see below).
 
 ### 2. Node Drill-down
 
-Breadcrumb: `Cluster > db-west-04`. Shows the node's workstreams in a table matching the per-node dashboard layout (STATE, NAME, NODE, TASK, TOKENS, CTX) with activity sub-lines. Includes a link to the node's own dashboard (`http://{server_url}/`).
+Breadcrumb: `Cluster > db-west-04`. Shows the node's workstreams in a table matching the per-node dashboard layout (STATE, NAME, MODEL, NODE, TASK, TOKENS, CTX) with activity sub-lines. Includes a link to the node's proxied server UI.
 
-**Deep linking:** Clicking a workstream row opens the node's server UI in a new tab with `?ws_id=<id>`, which auto-selects that workstream. A `↗` indicator appears on hover to signal the external navigation. Rows without a `server_url` are non-interactive.
+**Proxy deep-linking:** Clicking a workstream row opens the node's server UI in a new tab via the proxy at `/node/{node_id}/?ws_id=<id>`, which auto-selects that workstream. Users do not need direct network access to the server node.
 
 ### 3. Filtered Workstreams
 
-Breadcrumb: `Cluster > Running` or `Cluster > db-west-04`. Server-side paginated workstream table. NODE column values are clickable to filter further. Pagination controls at bottom. Workstream rows are deep-linkable when `server_url` is available (injected by the collector from the parent node).
+Breadcrumb: `Cluster > Running` or `Cluster > db-west-04`. Server-side paginated workstream table. NODE column values are clickable to filter further. Pagination controls at bottom. Workstream rows use proxy deep-links.
 
-All three views receive live updates via SSE — state cards update counts, node rows update metrics, workstream rows update state indicators.
+### 4. Workstream Creation Modal
+
+Triggered by the "+ new" header button. A modal dialog with:
+
+- **Node selector** — dropdown with three targeting modes: "Auto (best available)" picks the node with the most headroom, "General pool (any node)" pushes to the shared queue for any bridge to pick up, or a specific node from the list (showing capacity).
+- **Profile** — optional dropdown listing enabled workstream templates. Applies the template's model, auto-approve policy, token budget, and other behavioral settings at creation time.
+- **Name** — optional text input. Auto-generated if left empty.
+- **Model** — optional text input for a model alias from the target node's registry.
+
+On submit, `POST /v1/api/cluster/workstreams/new` dispatches the creation request. A toast confirms success; the SSE stream delivers the `ws_created` event to update the dashboard.
+
+All five views receive live updates via SSE — state cards update counts, node rows update metrics, workstream rows update state indicators.
+
+The browser maintains a local `clusterState` object that mirrors the cluster snapshot. It is initialized from the SSE `snapshot` event on connect (or via `GET /v1/api/cluster/snapshot` on initial page load) and updated incrementally by SSE events. View navigation reads from local state — no API round-trips needed after the initial snapshot.
+
+### 5. Admin Panel
+
+Accessed via the "admin" button in the header (visible when authenticated
+with `approve` scope). Provides user, API token, channel link, and workstream
+template management with 13 tabs (see also [Governance](governance.md) for
+the Roles, Policies, Templates, WS Templates, Usage, and Audit tabs, and
+[Settings](settings.md) for the database-backed configuration editor):
+
+**Users tab:**
+
+- Grid table listing all users (username, display name, role, creation date)
+- "Create User" button opens a modal with fields for username, display name,
+  and password (validated: username 1-64 ASCII, password min 8 characters)
+- Delete button on each row opens a styled confirmation modal before
+  removing the user and cascading to revoke all their tokens
+
+**Tokens tab:**
+
+- User selector dropdown to pick which user's tokens to manage
+- Grid table listing tokens for the selected user (name, prefix, scopes,
+  creation date)
+- Scope badges rendered as colored pills for visual clarity
+- "Create Token" button opens a modal with fields for token name and scope
+  checkboxes
+- On creation, a "Token Created" modal displays the raw `ts_`-prefixed
+  token with a copy button. The token is shown once and cannot be retrieved
+  again.
+- Revoke button on each row opens a styled confirmation modal before
+  deleting the token
+
+**Channels tab:**
+
+- User selector dropdown to pick which user's channel links to manage
+- Grid table listing linked channel accounts for the selected user
+  (channel type, channel user ID, creation date)
+- "Link Channel" button opens a modal with fields for channel type
+  (e.g. `discord`) and the platform user ID
+- Unlink button on each row opens a styled confirmation modal before
+  removing the channel mapping
+- Admins can force-link users who have not self-linked via `/link` in
+  Discord
+
+**Accessibility:**
+
+- Full keyboard navigation: focus traps in modals, Escape to close, arrow
+  keys for tab switching
+- Responsive layout with column hiding at 700px breakpoint
+
+**First-time setup:**
+
+The console also exposes `POST /v1/api/auth/setup` for first-time
+bootstrap. When no users exist, the setup wizard calls this public endpoint
+to create the initial admin user and receive a JWT in one step. See
+[Security: First-time setup](security.md#first-time-setup) for details.
+
+---
+
+## Scheduled Tasks
+
+The console includes a background **TaskScheduler** daemon that creates workstreams on a timed basis via the MQ broker. It supports cron-based recurring schedules and one-shot `at` schedules.
+
+### Architecture
+
+The scheduler runs as a daemon thread inside the console process. Every `check_interval` seconds (default 15) it:
+
+1. Acquires a distributed lock via Redis `SET NX EX` (prevents duplicate dispatch in multi-console deployments)
+2. Queries the storage backend for tasks whose `next_run <= now` and `enabled = true`
+3. Dispatches each due task as one or more `CreateWorkstreamMessage` via MQ
+4. Updates `last_run` and computes the next `next_run` (or disables one-shot `at` tasks)
+5. Releases the lock via Lua script (safe conditional delete)
+
+Run history is automatically pruned (runs older than 90 days) approximately once per hour.
+
+### Schedule Types
+
+| Type | Field | Behavior |
+|------|-------|----------|
+| `cron` | `cron_expr` | Recurring schedule using standard 5-field cron syntax. Requires `croniter`. |
+| `at` | `at_time` | One-shot: fires once at the given ISO 8601 timestamp (must include timezone), then auto-disables. |
+
+### Target Modes
+
+| Mode | Behavior |
+|------|----------|
+| `auto` | Picks the reachable node with the most available capacity |
+| `pool` | Pushes to the shared inbound queue (any bridge picks it up) |
+| `all` | Fan-out to all reachable nodes (capped at `max_fan_out`, default 20) |
+| `<node_id>` | Targets a specific node by ID |
+
+### Configuration
+
+| Parameter | Default | Description |
+|-----------|---------|-------------|
+| `check_interval` | `15.0` | Seconds between scheduler ticks |
+| `lock_ttl` | `60` | Distributed lock TTL in seconds |
+| `max_fan_out` | `20` | Maximum nodes for `all` target mode |
+
+Dependency: `croniter` (installed with turnstone).
+
+### Schedule API
+
+All schedule endpoints require `approve` scope. Maximum 200 schedules.
+
+#### `GET /v1/api/admin/schedules`
+
+List all scheduled tasks.
+
+```json
+{
+  "schedules": [
+    {
+      "task_id": "a1b2c3d4",
+      "name": "nightly-checks",
+      "description": "Run nightly health checks",
+      "schedule_type": "cron",
+      "cron_expr": "0 2 * * *",
+      "at_time": "",
+      "target_mode": "auto",
+      "model": "",
+      "initial_message": "Run the nightly health check suite.",
+      "auto_approve": false,
+      "auto_approve_tools": [],
+      "enabled": true,
+      "created_by": "u_admin",
+      "last_run": "2026-03-05T02:00:00Z",
+      "next_run": "2026-03-06T02:00:00Z",
+      "created": "2026-03-01T12:00:00Z",
+      "updated": "2026-03-05T02:00:01Z"
+    }
+  ]
+}
+```
+
+#### `POST /v1/api/admin/schedules`
+
+Create a scheduled task.
+
+Request:
+
+```json
+{
+  "name": "nightly-checks",
+  "description": "Run nightly health checks",
+  "schedule_type": "cron",
+  "cron_expr": "0 2 * * *",
+  "target_mode": "auto",
+  "initial_message": "Run the nightly health check suite.",
+  "auto_approve": false,
+  "enabled": true
+}
+```
+
+Required fields: `name`, `schedule_type`, `initial_message`. For `cron` schedules provide `cron_expr`; for `at` schedules provide `at_time` (ISO 8601 with timezone, must be in the future).
+
+Response: `ScheduleInfo` (same shape as list items above). Returns `400` for invalid cron syntax, naive timestamps, or past `at_time`. Returns `409` if the 200-schedule cap is reached.
+
+#### `GET /v1/api/admin/schedules/{task_id}`
+
+Get a single scheduled task. Returns `ScheduleInfo` or `404`.
+
+#### `PUT /v1/api/admin/schedules/{task_id}`
+
+Partial update — only include fields to change. If `schedule_type`, `cron_expr`, or `at_time` change, `next_run` is recomputed automatically.
+
+```json
+{
+  "enabled": false
+}
+```
+
+Response: updated `ScheduleInfo`. Returns `400` for validation errors, `404` if not found.
+
+#### `DELETE /v1/api/admin/schedules/{task_id}`
+
+Delete a scheduled task and all its run history. Returns `{"status": "ok"}` or `404`.
+
+#### `GET /v1/api/admin/schedules/{task_id}/runs?limit=50`
+
+List execution history for a task (most recent first). `limit` defaults to 50, max 200.
+
+```json
+{
+  "runs": [
+    {
+      "run_id": "r_abc123",
+      "task_id": "a1b2c3d4",
+      "node_id": "db-west-04",
+      "ws_id": "ws_xyz",
+      "correlation_id": "corr_789",
+      "started": "2026-03-05T02:00:00Z",
+      "status": "dispatched",
+      "error": ""
+    }
+  ]
+}
+```
+
+Status is `dispatched` on success or `failed` with an `error` message (e.g. no reachable nodes). Failed runs do not advance `next_run`.
 
 ---
 
@@ -201,6 +644,7 @@ CLI flags for `turnstone-console`:
 | `--redis-password` | `$REDIS_PASSWORD` | Redis password |
 | `--redis-db` | `0` | Redis DB |
 | `--poll-interval` | `10` | Node polling interval (seconds) |
+| `--auth-token` | `$TURNSTONE_AUTH_TOKEN` | Bearer token for server node communication and proxy |
 | `--log-level` | `INFO` | Log level |
 
 Config file (`~/.config/turnstone/config.toml`):
@@ -233,7 +677,7 @@ turnstone-server --port 8080
 turnstone-bridge --server-url http://localhost:8080 --node-id node-a
 
 # Start cluster console (one instance)
-turnstone-console --redis-host localhost --port 8090
+turnstone-console --redis-host localhost --port 8090 --auth-token "$TURNSTONE_AUTH_TOKEN"
 ```
 
-Open `http://localhost:8090` for the cluster dashboard.
+Open `http://localhost:8090` for the cluster dashboard. Create workstreams via the "+ new" button. Click any workstream to open the proxied server UI — no direct access to server ports required.

docs/diagrams/01-system-context.puml

@@ -9,7 +9,10 @@ actor "External Client\n(Python / CI)" as ext_client
 actor "Eval Harness" as eval_user
 
 ' External Systems
-cloud "LLM Provider\n(OpenAI-compatible API)" as llm
+cloud "LLM Providers" as llm {
+    component [OpenAI-compatible API\n(OpenAI, vLLM, llama.cpp)] as llm_openai
+    component [Anthropic Messages API] as llm_anthropic
+}
 database "Redis" as redis
 database "SQLite\n(.turnstone.db)" as sqlite
 
@@ -31,21 +34,21 @@ ext_client --> redis : Redis LIST\n(push commands)
 eval_user --> eval : Python API
 
 ' Internal connections
-cli --> llm : OpenAI Streaming API\n(HTTPS)
+cli --> llm : LLM Provider API\n(via provider adapters)
 cli --> sqlite : SQLite
 
-server --> llm : OpenAI Streaming API\n(HTTPS)
+server --> llm : LLM Provider API\n(via provider adapters)
 server --> sqlite : SQLite
 
-eval --> llm : OpenAI API\n(non-streaming)
+eval --> llm : LLM Provider API\n(non-streaming)
 eval --> sqlite : SQLite
 
-bridge --> server : HTTP REST\n(POST /api/send, etc.)
-bridge <-- server : SSE\n(GET /api/events)
+bridge --> server : HTTP REST\n(POST /v1/api/send, etc.)
+bridge <-- server : SSE\n(GET /v1/api/events)
 bridge --> redis : Redis LIST + PUBSUB\n+ STRING (routing, heartbeats)
 
-console --> redis : Redis PUBSUB + STRING\n(cluster channel, heartbeats)
-console --> server : HTTP polling\n(GET /api/dashboard)
+console --> redis : Redis PUBSUB + STRING + LIST\n(cluster events, heartbeats,\nworkstream creation commands)
+console --> server : HTTP polling + reverse proxy\n(GET /v1/api/dashboard,\nproxy /node/{id}/* traffic)
 
 sim --> redis : Redis LIST + PUBSUB\n+ STRING (heartbeats)
 

docs/diagrams/02-package-structure.puml

@@ -11,6 +11,8 @@ skinparam component {
     BackgroundColor<<console>> #B2EBF2
     BackgroundColor<<ui>> #F0F4C3
     BackgroundColor<<artifact>> #ECEFF1
+    BackgroundColor<<sdk>> #FFCDD2
+    BackgroundColor<<api>> #D1C4E9
 }
 
 ' Entry points
@@ -24,9 +26,11 @@ package "Entry Points" <<Rectangle>> {
 ' Core engine
 package "turnstone/core/" <<Rectangle>> {
     component [session.py\nChatSession, SessionUI] as session <<core>>
+    component [providers/\nLLMProvider, OpenAI, Anthropic] as providers <<core>>
     component [workstream.py\nWorkstreamManager] as workstream <<core>>
     component [tools.py\nTool loader] as tools <<core>>
-    component [memory.py\nSQLite + FTS5] as memory <<core>>
+    component [memory.py\nPersistence facade] as memory <<core>>
+    component [storage/\nStorageBackend protocol\nSQLite + PostgreSQL] as storage <<core>>
     component [metrics.py\nPrometheus metrics] as metrics <<core>>
     component [config.py\nTOML config] as config <<core>>
     component [safety.py\nPath validation] as safety <<core>>
@@ -36,7 +40,8 @@ package "turnstone/core/" <<Rectangle>> {
     component [auth.py\nAuthentication] as auth <<core>>
     component [healthcheck.py\nBackendHealthMonitor] as healthcheck <<core>>
     component [ratelimit.py\nRateLimiter] as ratelimit <<core>>
-    component [mcp_client.py\nMCPClientManager] as mcp <<core>>
+    component [mcp_client.py\nMCPClientManager\n(push + periodic refresh)] as mcp <<core>>
+    component [tool_search.py\nToolSearchManager, BM25] as toolsearch <<core>>
     component [model_registry.py\nModelRegistry] as registry <<core>>
 }
 
@@ -72,9 +77,26 @@ package "turnstone/ui/" <<Rectangle>> {
     component [spinner.py\nTerminal spinner] as spinner <<ui>>
 }
 
+' API schemas
+package "turnstone/api/" <<Rectangle>> {
+    component [schemas.py\nShared Pydantic models] as apischemas <<api>>
+    component [server_spec.py\nServer OpenAPI spec] as serverspec <<api>>
+    component [console_spec.py\nConsole OpenAPI spec] as consolespec <<api>>
+    component [openapi.py\nSpec builder] as openapi <<api>>
+    component [docs.py\nSwagger UI handler] as apidocs <<api>>
+}
+
+' SDK
+package "turnstone/sdk/" <<Rectangle>> {
+    component [server.py\nTurnstoneServer (sync+async)] as sdkserver <<sdk>>
+    component [console.py\nTurnstoneConsole (sync+async)] as sdkconsole <<sdk>>
+    component [events.py\n27 SSE event types] as sdkevents <<sdk>>
+    component [_base.py\nhttpx client base] as sdkbase <<sdk>>
+}
+
 ' Tool schemas
 package "turnstone/tools/" <<Rectangle>> {
-    component [*.json\n14 tool schemas] as schemas <<artifact>>
+    component [*.json\n18 tool schemas] as schemas <<artifact>>
 }
 
 ' Entry point dependencies
@@ -105,15 +127,19 @@ eval --> tools
 chat --> session
 
 ' Core internal deps
+session --> providers
 session --> tools
 session --> memory
+memory --> storage
 session --> safety
 session --> sandbox
 session --> edit
 session --> web
 session --> healthcheck
 session --> mcp : optional
+session --> toolsearch : optional
 session --> registry : optional
+registry --> providers
 healthcheck --> metrics
 mcp --> config
 registry --> config
@@ -149,4 +175,22 @@ consoleserver --> config
 consoleserver --> auth
 collector --> broker
 
+' API dependencies
+serverspec --> openapi
+consolespec --> openapi
+serverspec --> apischemas
+consolespec --> apischemas
+server --> apidocs
+server --> serverspec
+consoleserver --> apidocs
+consoleserver --> consolespec
+
+' SDK dependencies
+sdkserver --> sdkbase
+sdkconsole --> sdkbase
+sdkserver --> sdkevents
+sdkconsole --> sdkevents
+sdkserver --> apischemas : returns models
+sdkconsole --> apischemas : returns models
+
 @enduml

docs/diagrams/03-core-engine-classes.png

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b3f76042c8046560502fa3132351821d56e8526b13d38d7be8b839fcf3d5f648
-size 373463

docs/diagrams/03-core-engine-classes.puml

@@ -41,7 +41,7 @@ class "WorkstreamTerminalUI" as WsTermUI {
 }
 
 class "WebUI" as WebUI {
-    - _event_queue: Queue
+    - _listeners: list[Queue]
     - _approval_event: Event
     - _plan_event: Event
     - _ws_prompt_tokens: int
@@ -52,6 +52,8 @@ class "WebUI" as WebUI {
     Enqueues JSON events for SSE.
     Blocks on threading.Event for
     approval/plan review.
+    SSE handlers bridge Queue to
+    async via run_in_executor().
     --
     server.py
 }
@@ -63,15 +65,64 @@ class "NullUI" as NullUI {
     eval.py
 }
 
+' LLMProvider Protocol
+interface "LLMProvider" as LLMProvider <<Protocol>> {
+    + provider_name: str {property}
+    + get_capabilities(model) → ModelCapabilities
+    + create_streaming(client, model, messages, ...) → Iterator[StreamChunk]
+    + create_completion(client, model, messages, ...) → CompletionResult
+    + convert_tools(tools) → list[dict]
+    + retryable_error_names: frozenset[str] {property}
+    --
+    core/providers/_protocol.py
+}
+
+class "OpenAIProvider" as OpenAIProv {
+    Model capability lookup table
+    (GPT-5.x, O-series, search)
+    Passthrough: messages already
+    in OpenAI format.
+    Search models: web_search_options
+    + url_citation annotations.
+    --
+    core/providers/_openai.py
+}
+
+class "AnthropicProvider" as AnthropicProv {
+    Converts OpenAI messages to
+    Anthropic content blocks.
+    Adaptive + manual thinking.
+    Native web search via
+    web_search_20250305 server tool.
+    Lazy anthropic SDK import.
+    --
+    core/providers/_anthropic.py
+}
+
+' ModelCapabilities
+class "ModelCapabilities" as ModelCaps <<frozen>> {
+    + context_window: int
+    + max_output_tokens: int
+    + supports_temperature: bool
+    + token_param: str
+    + thinking_mode: str
+    + supports_effort: bool
+    + supports_web_search: bool
+    + supports_tool_search: bool
+    + supports_vision: bool
+}
+
 ' ChatSession
 class "ChatSession" as ChatSession {
-    - client: OpenAI
+    - client: Any
+    - provider: LLMProvider
     - model: str
     - ui: SessionUI
     - messages: list[dict]
     - _msg_tokens: list[int]
-    - _session_id: str
+    - _ws_id: str
     - _mcp_client: MCPClientManager | None
+    - _tool_search: ToolSearchManager | None
     - _registry: ModelRegistry | None
     + model_alias: str | None {property}
     - _tools: list[dict]
@@ -82,7 +133,7 @@ class "ChatSession" as ChatSession {
     --
     + send(user_input: str)
     + handle_command(command: str)
-    + resume_session(session_id: str)
+    + resume(ws_id: str)
     - _save_config()
     - _stream_response(stream) → dict
     - _create_stream_with_retry(msgs) → Stream (+ fallback)
@@ -91,6 +142,12 @@ class "ChatSession" as ChatSession {
     - _prepare_tool(tc) → item dict
     - _prepare_mcp_tool(call_id, name, args) → item dict
     - _exec_mcp_tool(item) → (call_id, output)
+    - _get_active_tools() → list[dict]
+    - _prepare_tool_search() → None
+    - _exec_tool_search(item) → (call_id, output)
+    - _on_mcp_tools_changed()
+    - _rebuild_tool_search()
+    + close()
     - _run_agent(messages, tools, ...) → str
     - _compact_messages(auto: bool)
     - _full_messages() → list[dict]
@@ -153,39 +210,77 @@ enum "WorkstreamState" as WsState {
 ' MCPClientManager
 class "MCPClientManager" as MCPMgr {
     - _sessions: dict[str, ClientSession]
+    - _per_server_tools: dict[str, list[dict]]
+    - _per_server_resources: dict[str, list[dict]]
+    - _per_server_prompts: dict[str, list[dict]]
     - _tools: list[dict]
     - _tool_map: dict[str, tuple]
+    - _resource_map: dict[str, tuple]
+    - _prompt_map: dict[str, tuple]
+    - _supports_list_changed: dict[str, bool]
+    - _listeners: list[Callable]
     --
     + start()
     + get_tools() → list[dict]
+    + get_resources() → list[dict]
+    + get_prompts() → list[dict]
     + is_mcp_tool(name) → bool
     + call_tool_sync(name, args) → str
+    + read_resource_sync(uri) → str
+    + get_prompt_sync(name, args?) → list[dict]
+    + refresh_sync(server?) → dict
+    + add_listener(callback)
+    + remove_listener(callback)
+    + server_names: list[str] {property}
     + shutdown()
     --
     Background asyncio event loop
     bridges async MCP SDK to
     sync ChatSession dispatch.
+    Push + periodic + manual refresh.
+    Resources + prompts discovered
+    alongside tools at startup.
     --
     core/mcp_client.py
 }
 
+' ToolSearchManager
+class "ToolSearchManager" as ToolSearchMgr {
+    - _all_tools: list[dict]
+    - _always_on: list[dict]
+    - _deferred: list[dict]
+    - _expanded: dict[str, None]
+    - _index: BM25Index
+    --
+    + should_activate() → bool
+    + get_visible_tools() → list[dict]
+    + get_deferred_tools() → list[dict]
+    + get_expanded_names() → list[str]
+    + search(query, k) → list[dict]
+    + expand_visible(names) → list[dict]
+    + get_search_tool_definition() → dict
+    + format_search_results(tools) → str
+}
+
 ' ModelRegistry
 class "ModelRegistry" as ModelReg {
     - _models: dict[str, ModelConfig]
-    - _clients: dict[str, OpenAI]
+    - _clients: dict[str, Any]
+    - _providers: dict[str, LLMProvider]
     - _client_lock: Lock
     + default: str
     + fallback: list[str]
     + agent_model: str | None
     --
     + resolve(alias) → (client, model, config)
-    + get_client(alias) → OpenAI
+    + get_client(alias) → Any
+    + get_provider(alias) → LLMProvider
     + has_alias(alias) → bool
     + list_aliases() → list[str]
     + shutdown()
     --
-    Thread-safe lazy client creation.
-    Loaded by load_model_registry()
+    Thread-safe lazy client + provider
+    creation. Loaded by load_model_registry()
     from CLI args + [models.*] config.
     --
     core/model_registry.py
@@ -193,6 +288,7 @@ class "ModelRegistry" as ModelReg {
 
 class "ModelConfig" as ModelCfg <<frozen>> {
     + alias: str
+    + provider: str
     + base_url: str
     + model: str
     + context_window: int
@@ -260,8 +356,13 @@ TerminalUI <|-- WsTermUI
 SessionUI <|.. WebUI
 SessionUI <|.. NullUI
 
+LLMProvider <|.. OpenAIProv
+LLMProvider <|.. AnthropicProv
+
 ChatSession --> SessionUI : uses
+ChatSession --> LLMProvider : delegates LLM calls
 ChatSession --> MCPMgr : optional
+ChatSession --o ToolSearchMgr : _tool_search
 ChatSession --> ModelReg : optional
 ChatSession <|-- HeadlessSession
 
@@ -273,6 +374,8 @@ Ws --> "1" WsState : has
 WsMgr ..> ChatSession : creates via\nsession_factory(ui, model_alias)
 
 ModelReg --> "*" ModelCfg : holds
+ModelReg --> "*" LLMProvider : caches
+LLMProvider --> ModelCaps : returns
 
 ChatSession --> HealthMon : checks circuit
 HealthMon --> "1" CircuitState : has
@@ -282,6 +385,8 @@ note bottom of ChatSession
   Central engine: multi-turn LLM loop
   with tool dispatch, agent sub-sessions,
   context compaction, and memory persistence.
+  Provider-agnostic — delegates all LLM
+  communication to LLMProvider adapters.
 
   core/session.py (~2700 lines)
 end note

docs/diagrams/04-conversation-turn.png

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fa1c94a0a7489cb9e6e6cad17f7e3577c458ec89d2a47fb2669fe935768837cf
-size 276863

docs/diagrams/04-conversation-turn.puml

@@ -8,7 +8,7 @@ skinparam sequenceLifeLineBackgroundColor #F5F5F5
 participant "User /\nHTTP Client" as User
 participant "ChatSession" as CS
 participant "SessionUI" as UI
-participant "OpenAI API\n(LLM)" as LLM
+participant "LLMProvider\n(OpenAI / Anthropic)" as LLM
 participant "Tool Executor\n(ThreadPool)" as TP
 database "SQLite" as DB
 
@@ -18,7 +18,7 @@ User -> CS : send(user_input)
 activate CS
 
 CS -> CS : messages.append({role: "user", content: input})
-CS -> DB : save_message(session_id, "user", input)
+CS -> DB : save_message(ws_id, "user", input)
 
 == LLM Call Loop ==
 
@@ -27,7 +27,7 @@ group loop [while tool_calls present]
     CS -> UI : on_state_change("thinking")
     CS -> UI : on_thinking_start()
 
-    CS -> LLM : client.chat.completions.create(\n  model, messages, tools,\n  stream=True, stream_options={include_usage})
+    CS -> LLM : provider.create_streaming(\n  client, model, messages, tools, ...)\n  (normalized StreamChunk iterator)
     activate LLM
 
     note right of CS
@@ -52,9 +52,19 @@ group loop [while tool_calls present]
             CS -> UI : on_content_token(text)
         else tool_call delta
             CS -> CS : accumulate in tool_calls_acc
+        else info_delta present
+            CS -> UI : on_info(text)\n(e.g. server-side web search status)
         end
     end
 
+        note right of CS
+          **Cancellation checkpoint:**
+          _check_cancelled() runs per chunk.
+          If cancel_event is set, raises
+          GenerationCancelled — preserves
+          partial content, emits idle state.
+        end note
+
     LLM --> CS : stream complete (usage stats)
     deactivate LLM
 
@@ -63,8 +73,8 @@ group loop [while tool_calls present]
 
     CS -> CS : _update_token_table()\ncalibrate chars_per_token ratio
     CS -> CS : messages.append(assistant_msg)
-    CS -> DB : save_message(session_id, "assistant", content)
-    CS -> DB : save_message(session_id, "tool_call", ...) ×N
+    CS -> DB : save_message(ws_id, "assistant", content)
+    CS -> DB : save_message(ws_id, "tool_call", ...) ×N
 
     == Tool Dispatch (if tool_calls) ==
 
@@ -110,14 +120,14 @@ group loop [while tool_calls present]
         note right of TP
           Parallel execution:
           bash → Popen + line-by-line streaming
-          read_file → open().read()
+          read_file → open().read() or base64 image
           search → grep subprocess
           edit_file → string replace
           task/plan → _run_agent() sub-loop
           math → sandboxed subprocess
           web_fetch → httpx + LLM summarize
-          web_search → Tavily API
-          remember/recall/forget → SQLite
+          web_search → provider-native or Tavily fallback
+          memory/recall → SQLite
         end note
 
         note right of TP
@@ -134,7 +144,7 @@ group loop [while tool_calls present]
 
         loop for each result
             CS -> CS : messages.append({role: "tool", ...})
-            CS -> DB : save_message(session_id, "tool_result", ...)
+            CS -> DB : save_message(ws_id, "tool_result", ...)
         end
 
         opt user_feedback from approval
@@ -142,6 +152,12 @@ group loop [while tool_calls present]
         end
 
         note right of CS : Loop back for next LLM call
+
+    else GenerationCancelled
+        CS -> CS : Preserve partial content\nor roll back incomplete tools
+        CS -> UI : on_info("[Generation cancelled]")
+        CS -> UI : on_state_change("idle")
+        CS --> User : return (no re-raise)
     end
 
 end

docs/diagrams/05-tool-pipeline.png

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:06fb9faa29c11395c6fc54ebddc79994b000dee78d56e0c13cb689fd6a82e37a
-size 237255

docs/diagrams/05-tool-pipeline.puml

@@ -24,27 +24,30 @@ partition "Phase 1: Prepare" #E8F5E9 {
         :Dispatch to _prepare_{func_name}();
 
         note right
-            **Dispatch table (14 tools):**
-            ┌─────────────┬──────────────────┐
-            │ Tool        │ Needs Approval?  │
-            ├─────────────┼──────────────────┤
-            │ bash        │ ✓ Yes            │
-            │ read_file   │ ✗ Auto-approve   │
-            │ write_file  │ ✓ Yes            │
-            │ edit_file   │ ✓ Yes            │
-            │ search      │ ✗ Auto-approve   │
-            │ math        │ ✓ Yes            │
-            │ man         │ ✗ Auto-approve   │
-            │ web_fetch   │ ✓ Yes            │
-            │ web_search  │ ✓ Yes            │
-            │ task        │ ✓ Yes            │
-            │ plan        │ ✓ Yes            │
-            │ remember    │ ✗ Auto-approve   │
-            │ recall      │ ✗ Auto-approve   │
-            │ forget      │ ✗ Auto-approve   │
-            ├─────────────┼──────────────────┤
-            │ mcp__*      │ ✓ Yes (external) │
-            └─────────────┴──────────────────┘
+            **Dispatch table (17 tools):**
+            ┌───────────────┬──────────────────┐
+            │ Tool          │ Needs Approval?  │
+            ├───────────────┼──────────────────┤
+            │ bash          │ ✓ Yes            │
+            │ read_file     │ ✗ Auto-approve   │
+            │ write_file    │ ✓ Yes            │
+            │ edit_file     │ ✓ Yes            │
+            │ search        │ ✗ Auto-approve   │
+            │ math          │ ✗ Auto-approve   │
+            │ man           │ ✗ Auto-approve   │
+            │ web_fetch     │ ✗ Auto-approve   │
+            │ web_search    │ ✗ Auto-approve   │
+            │ tool_search   │ ✗ Auto-approve   │
+            │ task          │ ✓ Yes            │
+            │ plan          │ ✓ Yes            │
+            │ memory        │ ✗ Auto-approve   │
+            │ recall        │ ✗ Auto-approve   │
+            │ notify        │ ✗ Auto-approve   │
+            │ read_resource │ ✓ Yes            │
+            │ use_prompt    │ ✓ Yes            │
+            ├───────────────┼──────────────────┤
+            │ mcp__*        │ ✓ Yes (external) │
+            └───────────────┴──────────────────┘
         end note
 
         :Build item dict:
@@ -86,6 +89,8 @@ partition "Phase 2: Approve" #FFF3E0 {
 }
 
 partition "Phase 3: Execute" #E3F2FD {
+    :_check_cancelled();
+    note right: Cancellation checkpoint:\nraises GenerationCancelled if\ncancel event is set
     if (single tool call?) then (yes)
         :Execute sequentially:\nrun_one(items[0]);
     else (multiple)
@@ -98,19 +103,22 @@ partition "Phase 3: Execute" #E3F2FD {
         if item.denied → return denial message
         else → item["execute"](item)
           ├─ _exec_bash: subprocess.run(["bash", script.sh])
-          ├─ _exec_read_file: open().readlines()
+          ├─ _exec_read_file: open().readlines() or _exec_read_image (base64)
           ├─ _exec_write_file: makedirs + write
           ├─ _exec_edit_file: find_occurrences + replace
           ├─ _exec_search: grep subprocess
           ├─ _exec_math: sandboxed subprocess
           ├─ _exec_man: man/info subprocess
           ├─ _exec_web_fetch: httpx.get + LLM summary
-          ├─ _exec_web_search: Tavily API POST
+          ├─ _exec_web_search: Tavily API POST (fallback for local models)
+          ├─ _exec_tool_search: BM25 search + expand_visible()
           ├─ _exec_task: _run_agent(TASK_AGENT_TOOLS)
           ├─ _exec_plan: _run_agent(AGENT_TOOLS, read-only)
-          ├─ _exec_remember: SQLite INSERT OR REPLACE
-          ├─ _exec_recall: SQLite FTS5/LIKE search
-          ├─ _exec_forget: SQLite DELETE
+          ├─ _exec_notify: HTTP POST to channel gateway
+          ├─ _exec_memory: structured memory save/search/delete/list
+          ├─ _exec_recall: conversation history FTS5 search
+          ├─ _exec_read_resource: MCPClientManager.read_resource_sync()
+          ├─ _exec_use_prompt: MCPClientManager.get_prompt_sync()
           └─ _exec_mcp_tool: MCPClientManager.call_tool_sync()
     end note
 

docs/diagrams/06-mq-protocol.png

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:13b2da77312e5abb44c81aabb7f0addccab31d9bc7e8ef2f1c3563ef985ed503
-size 186941

docs/diagrams/06-mq-protocol.puml

@@ -58,6 +58,9 @@ package "Inbound Messages (Client → Bridge)" #FFF3E0 {
         + auto_approve: bool = False
         + auto_approve_tools: list[str] = []
         + target_node: str = ""
+        + initial_message: str = ""
+        + template: str = ""
+        + ws_template: str = ""
     }
 
     class CloseWorkstreamMessage {
@@ -78,6 +81,12 @@ package "Inbound Messages (Client → Bridge)" #FFF3E0 {
         type = "list_nodes"
     }
 
+    class CancelMessage {
+        type = "cancel"
+        --
+        + ws_id: str
+    }
+
     IM <|-- SendMessage
     IM <|-- ApproveMessage
     IM <|-- PlanFeedbackMessage
@@ -87,6 +96,7 @@ package "Inbound Messages (Client → Bridge)" #FFF3E0 {
     IM <|-- ListWorkstreamsMessage
     IM <|-- HealthMessage
     IM <|-- ListNodesMessage
+    IM <|-- CancelMessage
 }
 
 package "Outbound Events (Bridge → Client)" #E3F2FD {

docs/diagrams/07-message-routing.puml

@@ -18,20 +18,20 @@ note right of Redis : Shared queue — any bridge can pick up
 BridgeA -> Redis : BLPOP [turnstone:inbound:nodeA,\n         turnstone:inbound]
 Redis --> BridgeA : SendMessage (from shared queue)
 
-BridgeA -> ServerA : POST /api/workstreams/new\n{name:"", auto_approve:false}
+BridgeA -> ServerA : POST /v1/api/workstreams/new\n{name:"", auto_approve:false}
 ServerA --> BridgeA : {ws_id:"abc12345", name:"ws-abc1"}
 
 BridgeA -> Redis : SET turnstone:ws:abc12345 "nodeA"
 note right : Register workstream ownership
 
-BridgeA -> ServerA : GET /api/events?ws_id=abc12345
+BridgeA -> ServerA : GET /v1/api/events?ws_id=abc12345
 note right : Start per-WS SSE thread
 
 BridgeA -> Redis : PUBLISH turnstone:events:global\nWorkstreamCreatedEvent
 
 BridgeA -> Redis : PUBLISH turnstone:events:cluster\nClusterStateEvent(ws_id, state:"idle", node_id:"nodeA")
 
-BridgeA -> ServerA : POST /api/send\n{message:"...", ws_id:"abc12345"}
+BridgeA -> ServerA : POST /v1/api/send\n{message:"...", ws_id:"abc12345"}
 ServerA --> BridgeA : {status:"ok"}
 
 BridgeA -> Redis : PUBLISH turnstone:events:abc12345\nAckEvent(status:"ok")
@@ -79,7 +79,7 @@ note right of BridgeA
   1. _ws_auto_approve[ws_id]? → auto
   2. All tools in safe set? → auto
      (read_file, search, man,
-      remember, recall, forget)
+      memory, recall)
   3. Otherwise → manual approval
 end note
 
@@ -93,7 +93,7 @@ note right : Response queue — bypasses inbound queue
 BridgeA -> Redis : BLPOP turnstone:resp:req_xyz\n(spawned approval thread, timeout 300s)
 Redis --> BridgeA : ApproveMessage
 
-BridgeA -> ServerA : POST /api/approve\n{approved:true, ws_id:"abc12345"}
+BridgeA -> ServerA : POST /v1/api/approve\n{approved:true, ws_id:"abc12345"}
 
 == Heartbeat (continuous) ==
 

docs/diagrams/09-workstream-states.puml

@@ -40,6 +40,12 @@ running --> error : Exception during\ntool execution
 
 error --> thinking : New send() call\n_emit_state("thinking")
 
+thinking --> idle : cancel() called\n_emit_state("idle")
+
+running --> idle : cancel() called\n_emit_state("idle")
+
+attention --> idle : cancel() unblocks\napproval/plan wait\n_emit_state("idle")
+
 note right of thinking
   **Emitted via:**
   session._emit_state(state)

docs/diagrams/11-console-data-flow.puml

@@ -1,13 +1,14 @@
 @startuml
 !theme plain
-title Turnstone — Console Dashboard Data Collection
+title Turnstone — Console Dashboard Data Flow
 
 skinparam sequenceArrowThickness 1.5
 
 participant "Browser" as Browser
-participant "Console\nHTTP Server" as Server
+participant "Console\nStarlette App" as Server
 participant "ClusterCollector" as CC
 collections "Redis" as Redis
+participant "Node-A Bridge" as BridgeA
 participant "Node-A\n(real server)" as NodeA
 participant "Node-B\n(sim node)" as NodeB
 
@@ -67,17 +68,29 @@ note right of CC
   from the cluster event channel.
 end note
 
-CC -> NodeA : GET /api/dashboard
+CC -> NodeA : GET /v1/api/dashboard
 activate NodeA
 NodeA --> CC : {workstreams: [...],\naggregate: {total_tokens, ...}}
 deactivate NodeA
 
 CC -> NodeA : GET /health
 activate NodeA
-NodeA --> CC : {status:"ok", model:"...",\nworkstreams:{total, idle, ...}}
+NodeA --> CC : {status:"ok", version:"0.3.0",\nmodel:"...", workstreams:{...}}
 deactivate NodeA
 
+CC -> CC : Diff old vs new workstream IDs
 CC -> CC : Replace NodeSnapshot["nodeA"]\n.workstreams, .health, .aggregate
+CC -> CC : _fanout(ws_created) for\nnewly appeared workstreams
+CC -> CC : _fanout(ws_closed) for\nremoved workstreams
+
+note right of CC
+  Poll-diff fanout ensures
+  browser SSE clients learn
+  about workstreams that
+  appeared without a real-time
+  cluster event (e.g. bridge
+  startup recovery).
+end note
 
 CC -x NodeB : (SKIPPED: sim:// URL)
 
@@ -85,40 +98,148 @@ deactivate CC
 
 == Browser SSE Stream ==
 
-Browser -> Server : GET /api/cluster/events
+Browser -> Server : GET /v1/api/cluster/events
 activate Server
 
-Server -> CC : register_listener(queue)
-note right : Per-client queue.Queue(maxsize=500)
+Server -> CC : get_snapshot()
+CC --> Server : ClusterSnapshot\n(full current state)
 
-loop continuous
+Server -> CC : register_listener(queue)
+note right : Per-client queue.Queue(maxsize=500)\nSSE via EventSourceResponse + run_in_executor()
+
+Server -> Browser : data: {"type":"snapshot",...}\n(full state as first SSE event)
+
+loop continuous (incremental updates)
     CC -> Server : event via listener queue\n(from any of the 3 threads)
     Server -> Browser : data: {"type":"cluster_state",...}\n\n
 end
 
-alt timeout (5s no events)
-    Server -> Browser : : keepalive\n\n
+alt keepalive (sse-starlette ping=5)
+    Server -> Browser : : ping\n\n
 end
 
 Browser -> Server : connection closed
 Server -> CC : unregister_listener(queue)
 deactivate Server
 
-== Browser REST Requests ==
+== Browser REST: Snapshot ==
 
-Browser -> Server : GET /api/cluster/overview
-Server -> CC : get_overview()
-CC --> Server : {nodes: 10, workstreams: 47,\nstates: {running:5, thinking:3, ...},\naggregate: {total_tokens: 50000}}
+Browser -> Server : GET /v1/api/cluster/snapshot
+Server -> CC : get_snapshot()
+CC --> Server : ClusterSnapshot\n(full current state)
 Server --> Browser : JSON response
 
-Browser -> Server : GET /api/cluster/nodes?sort=activity
+== Browser REST Requests ==
+
+Browser -> Server : GET /v1/api/cluster/overview
+Server -> CC : get_overview()
+CC --> Server : {nodes: 10, workstreams: 47,\nstates: {running:5, ...},\naggregate: {total_tokens: 50000},\nversion_drift: false, versions: ["0.3.0"]}
+Server --> Browser : JSON response
+
+Browser -> Server : GET /v1/api/cluster/nodes?sort=activity
 Server -> CC : get_nodes(sort_by="activity")
 CC --> Server : {nodes: [...], total: 10}
 Server --> Browser : JSON response
 
-Browser -> Server : GET /api/cluster/workstreams\n?state=running&node=sim-0003
+Browser -> Server : GET /v1/api/cluster/workstreams\n?state=running&node=sim-0003
 Server -> CC : get_workstreams(state="running",\nnode="sim-0003")
 CC --> Server : {workstreams: [...], total: 5,\npage: 1, per_page: 50, pages: 1}
 Server --> Browser : JSON response
 
+== Workstream Creation (via MQ) ==
+
+Browser -> Server : POST /v1/api/cluster/workstreams/new\n{node_id:"nodeA", name:"new-task"}
+activate Server #FFECB3
+
+Server -> CC : _pick_best_node() or\nget_node_detail(node_id)
+CC --> Server : node validated
+
+Server -> Server : Build CreateWorkstreamMessage\n{target_node:"nodeA", name:"new-task"}
+
+Server -> Redis : RPUSH turnstone:inbound:nodeA\n(directed queue)
+Server --> Browser : {status:"ok", correlation_id:"abc",\ntarget_node:"nodeA"}
+deactivate Server
+
+note right of Redis
+  Bridge on Node-A picks up the
+  message from its directed queue,
+  POSTs to /v1/api/workstreams/new,
+  registers ownership, publishes
+  ws_created to cluster channel.
+end note
+
+Redis --> BridgeA : BLPOP turnstone:inbound:nodeA
+activate BridgeA
+BridgeA -> NodeA : POST /v1/api/workstreams/new\n{name:"new-task"}
+NodeA --> BridgeA : {ws_id:"ws789", name:"new-task"}
+BridgeA -> Redis : SET turnstone:ws:ws789 = nodeA
+BridgeA -> Redis : PUBLISH turnstone:events:cluster\n{type:"ws_created", ws_id:"ws789",\nnode_id:"nodeA", name:"new-task"}
+deactivate BridgeA
+
+Redis --> CC : ws_created event
+CC -> CC : Add workstream to\nNodeSnapshot["nodeA"]
+CC -> CC : _fanout(event)
+Server -> Browser : SSE: data: {"type":"ws_created",...}
+
+== Reverse Proxy (server UI through console port) ==
+
+Browser -> Server : GET /node/nodeA/
+activate Server #FFF9C4
+
+Server -> CC : get_node_detail("nodeA")\n→ server_url = "http://10.0.1.1:8080"
+
+Server -> NodeA : GET http://10.0.1.1:8080/\n(via httpx.AsyncClient)
+activate NodeA
+NodeA --> Server : index.html
+deactivate NodeA
+
+Server -> Server : Rewrite static paths:\nhref="/static/" → "/node/nodeA/static/"\nInject console-return banner\nafter <body>
+
+Server --> Browser : Rewritten HTML
+deactivate Server
+
+Browser -> Server : GET /node/nodeA/static/app.js
+activate Server #FFF9C4
+
+Server -> NodeA : GET http://10.0.1.1:8080/static/app.js
+activate NodeA
+NodeA --> Server : app.js
+deactivate NodeA
+
+Server -> Server : Prepend JS proxy shim:\nOverride fetch() and EventSource()\nto prepend "/node/nodeA" prefix
+
+Server --> Browser : Shimmed app.js
+deactivate Server
+
+note right of Browser
+  All fetch("/v1/api/send") calls in the
+  server UI now become fetch("/node/nodeA/v1/api/send"),
+  routed through the console proxy.
+end note
+
+Browser -> Server : GET /node/nodeA/v1/api/events?ws_id=ws789
+activate Server #FFF9C4
+
+Server -> NodeA : GET http://10.0.1.1:8080/v1/api/events?ws_id=ws789\n(SSE stream via httpx.AsyncClient timeout=None)
+activate NodeA
+
+loop SSE streaming
+    NodeA --> Server : data: {"type":"content","text":"..."}\n\n
+    Server --> Browser : data: {"type":"content","text":"..."}\n\n
+end
+
+deactivate NodeA
+deactivate Server
+
+Browser -> Server : POST /node/nodeA/v1/api/send\n{message:"hello", ws_id:"ws789"}
+activate Server #FFF9C4
+
+Server -> NodeA : POST http://10.0.1.1:8080/v1/api/send\n(body forwarded)
+activate NodeA
+NodeA --> Server : {status:"ok"}
+deactivate NodeA
+
+Server --> Browser : {status:"ok"}
+deactivate Server
+
 @enduml

docs/diagrams/12-deployment.puml

@@ -83,12 +83,12 @@ mqclient --> redis : Redis protocol\nport 6379
 server --> redis : Redis protocol\n(6379)
 server --> llm_api : OpenAI API\n(HTTPS/HTTP)
 
-bridge --> server : HTTP REST\n(POST /api/send, etc.)
-bridge <-- server : SSE\n(GET /api/events)
+bridge --> server : HTTP REST\n(POST /v1/api/send, etc.)
+bridge <-- server : SSE\n(GET /v1/api/events)
 bridge --> redis : Redis protocol\n(queues + pubsub)
 
-console --> redis : Redis PUBSUB\n(cluster channel)
-console --> server : HTTP polling\n(GET /api/dashboard)
+console --> redis : Redis PUBSUB + LIST\n(cluster events,\nws creation commands)
+console --> server : HTTP polling + proxy\n(GET /v1/api/dashboard,\nproxy /node/{id}/*)
 
 sim --> redis : Redis protocol\n(queues + pubsub + keys)
 

docs/diagrams/13-sdk-architecture.puml

@@ -0,0 +1,158 @@
+@startuml
+!theme plain
+title Turnstone — Client SDK Architecture
+
+skinparam class {
+    BackgroundColor<<async>> #C8E6C9
+    BackgroundColor<<sync>> #B8D4E3
+    BackgroundColor<<event>> #FFE0B2
+    BackgroundColor<<type>> #F0F4C3
+    BackgroundColor<<ts>> #E1BEE7
+}
+
+skinparam packageBorderColor #888888
+skinparam ArrowColor #555555
+
+' Python SDK
+package "turnstone/sdk/ (Python)" {
+    abstract class _BaseClient <<async>> {
+        - _client: httpx.AsyncClient
+        - _owns_client: bool
+        + _request(method, path, ...) → T
+        + _stream_sse(path, ...) → AsyncIterator
+        + aclose()
+    }
+
+    class AsyncTurnstoneServer <<async>> {
+        + list_workstreams()
+        + dashboard()
+        + create_workstream()
+        + close_workstream()
+        + send(message, ws_id)
+        + approve()
+        + plan_feedback()
+        + command()
+        + cancel(ws_id)
+        + stream_events(ws_id)
+        + stream_global_events()
+        + send_and_wait()
+        + list_saved_workstreams()
+        + login() / logout()
+        + health()
+    }
+
+    class AsyncTurnstoneConsole <<async>> {
+        + overview()
+        + nodes()
+        + workstreams()
+        + node_detail()
+        + snapshot()
+        + create_workstream()
+        + stream_cluster_events()
+        + login() / logout()
+        + health()
+    }
+
+    class TurnstoneServer <<sync>> {
+        - _async: AsyncTurnstoneServer
+        - _runner: _SyncRunner
+        .. delegates all methods ..
+        + __enter__ / __exit__
+    }
+
+    class TurnstoneConsole <<sync>> {
+        - _async: AsyncTurnstoneConsole
+        - _runner: _SyncRunner
+        .. delegates all methods ..
+        + __enter__ / __exit__
+    }
+
+    class _SyncRunner <<sync>> {
+        - _loop: EventLoop
+        - _thread: Thread
+        + run(coro) → T
+        + run_iter(async_gen) → Iterator
+        + close()
+    }
+
+    class TurnResult <<type>> {
+        + ws_id: str
+        + content_parts: list[str]
+        + reasoning_parts: list[str]
+        + tool_results: list
+        + errors: list[str]
+        + timed_out: bool
+        --
+        + content: str
+        + reasoning: str
+        + ok: bool
+    }
+
+    class ServerEvent <<event>> {
+        + type: str
+        + ws_id: str
+        + from_dict() → ServerEvent
+    }
+
+    class ClusterEvent <<event>> {
+        + type: str
+        + from_dict() → ClusterEvent
+    }
+
+    _BaseClient <|-- AsyncTurnstoneServer
+    _BaseClient <|-- AsyncTurnstoneConsole
+    TurnstoneServer --> AsyncTurnstoneServer : wraps
+    TurnstoneServer --> _SyncRunner : uses
+    TurnstoneConsole --> AsyncTurnstoneConsole : wraps
+    TurnstoneConsole --> _SyncRunner : uses
+    AsyncTurnstoneServer ..> TurnResult : returns
+    AsyncTurnstoneServer ..> ServerEvent : yields
+    AsyncTurnstoneConsole ..> ClusterEvent : yields
+}
+
+' TypeScript SDK
+package "sdk/typescript/ (TypeScript)" {
+    class "BaseClient" as TSBase <<ts>> {
+        # baseUrl: string
+        # token: string
+        # fetchFn: fetch
+        # request<T>()
+        # streamSSE<T>()
+    }
+
+    class "TurnstoneServer" as TSServer <<ts>> {
+        + listWorkstreams()
+        + send()
+        + streamEvents()
+        + sendAndWait()
+        ...
+    }
+
+    class "TurnstoneConsole" as TSConsole <<ts>> {
+        + overview()
+        + nodes()
+        + snapshot()
+        + clusterEvents()
+        ...
+    }
+
+    TSBase <|-- TSServer
+    TSBase <|-- TSConsole
+}
+
+' External connections
+class "turnstone-server :8080" as Server <<artifact>>
+class "turnstone-console :8081" as Console <<artifact>>
+
+AsyncTurnstoneServer --> Server : httpx REST + SSE
+AsyncTurnstoneConsole --> Console : httpx REST + SSE
+TSServer --> Server : fetch REST + SSE
+TSConsole --> Console : fetch REST + SSE
+
+note right of AsyncTurnstoneServer
+  Returns Pydantic models from
+  turnstone.api.server_schemas
+  (no type duplication)
+end note
+
+@enduml

docs/diagrams/14-storage-architecture.puml

@@ -0,0 +1,171 @@
+@startuml
+!theme plain
+title Turnstone — Storage Architecture
+
+skinparam class {
+    BackgroundColor<<protocol>> #E8EAF6
+    BackgroundColor<<sqlite>> #C8E6C9
+    BackgroundColor<<postgres>> #B3E5FC
+    BackgroundColor<<facade>> #FFF9C4
+    BackgroundColor<<migration>> #FFE0B2
+    BackgroundColor<<schema>> #F3E5F5
+}
+
+' -- Protocol --
+interface "StorageBackend" as SB <<protocol>> {
+    +save_message(ws_id, role, content, ...)
+    +load_messages(ws_id) → list[dict]
+    +register_workstream(ws_id, node_id, name, state)
+    +update_workstream_state(ws_id, state)
+    +update_workstream_name(ws_id, name)
+    +set_workstream_alias(ws_id, alias) → bool
+    +update_workstream_title(ws_id, title)
+    +resolve_workstream(alias_or_id) → str | None
+    +delete_workstream(ws_id) → bool
+    +prune_workstreams(retention_days) → (int, int)
+    +list_workstreams(node_id, limit) → list
+    +save_workstream_config(ws_id, config)
+    +load_workstream_config(ws_id) → dict
+    +kv_get(key) → str | None
+    +kv_set(key, value) → str | None
+    +kv_delete(key) → bool
+    +kv_list() → list[(str, str)]
+    +kv_search(query) → list[(str, str)]
+    +search_history(query, limit) → list
+    +search_history_recent(limit) → list
+    +create_user(user_id, username, display_name, pw_hash)
+    +get_user(user_id) / get_user_by_username(username)
+    +list_users() / delete_user(user_id)
+    +create_api_token(...) / get_api_token_by_hash(hash)
+    +list_api_tokens(user_id) / delete_api_token(id)
+    +close()
+}
+
+' -- Backends --
+class "SQLiteBackend" as SQLite <<sqlite>> {
+    -_engine: sa.Engine
+    -_fts5_available: bool
+    +__init__(path: str)
+    --
+    FTS5 full-text search
+    Default pool, check_same_thread=False
+}
+
+class "PostgreSQLBackend" as PG <<postgres>> {
+    -_engine: sa.Engine
+    +__init__(url: str, pool_size: int)
+    --
+    tsvector + ILIKE search
+    Connection pooling
+}
+
+' -- Schema --
+class "_schema.py" as Schema <<schema>> {
+    +metadata: MetaData
+    +memories: Table
+    +conversations: Table
+    +workstreams: Table (node_id, alias, title,\n    state, ws_template_id, ws_template_version)
+    +workstream_config: Table
+    +users: Table (username, password_hash)
+    +api_tokens: Table (token_hash, scopes)
+    +channel_users: Table (channel_type)
+    +workstream_templates: Table (name, model,\n    system_prompt, token_budget, version)
+    +workstream_template_versions: Table\n    (template_id, version, snapshot)
+    +scheduled_tasks: Table (..., ws_template)
+    --
+    SQLAlchemy Core
+    Single source of truth
+}
+
+' -- Migration --
+class "_migrate.py" as Migrate <<migration>> {
+    +run_migrations(storage, backend)
+    -_bootstrap_existing_sqlite()
+    --
+    Programmatic Alembic
+    Auto-bootstrap existing DBs
+}
+
+class "migrations/" as Versions <<migration>> {
+    001_initial_schema.py
+    002_user_identity.py
+}
+
+' -- Registry --
+class "_registry.py" as Registry {
+    -_storage: StorageBackend | None
+    +init_storage(backend, path, url) → StorageBackend
+    +get_storage() → StorageBackend
+    +reset_storage()
+    --
+    Auto-initializes SQLite
+    if not configured
+}
+
+' -- Facade --
+class "memory.py" as Facade <<facade>> {
+    +save_message()
+    +load_messages()
+    +register_workstream()
+    +update_workstream_state()
+    +save_workstream_config()
+    +save_memory() / delete_memory()
+    +search_memories()
+    +... (all delegated functions)
+    --
+    Thin delegation to
+    get_storage()
+    Silent failure behavior
+}
+
+' -- Consumers --
+class "session.py\nChatSession" as Session {
+}
+
+class "server.py\nWeb UI" as Server {
+}
+
+class "cli.py\nTerminal" as CLI {
+}
+
+' -- Relationships --
+SQLite ..|> SB
+PG ..|> SB
+
+SQLite --> Schema : uses
+PG --> Schema : uses
+
+Registry --> SB : creates
+Registry --> Migrate : calls
+
+Migrate --> Versions : applies
+Migrate --> Schema : references
+
+Facade --> Registry : get_storage()
+
+Session --> Facade : imports
+Server --> Facade : imports
+CLI --> Facade : imports
+
+' -- Config --
+note right of Registry
+  [database]
+  backend = "sqlite" | "postgresql"
+  url = "postgresql+psycopg://..."
+  path = ".turnstone.db"
+  pool_size = 5
+end note
+
+note bottom of SQLite
+  Default backend.
+  Zero-config for
+  single-node / dev.
+end note
+
+note bottom of PG
+  Production backend.
+  Multi-node / Docker
+  default.
+end note
+
+@enduml

docs/diagrams/15-auth-architecture.puml

@@ -0,0 +1,179 @@
+@startuml
+!theme plain
+title Turnstone — Authentication Architecture
+
+skinparam class {
+    BackgroundColor<<core>> #E8EAF6
+    BackgroundColor<<jwt>> #C8E6C9
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<endpoint>> #FFE0B2
+    BackgroundColor<<scope>> #F3E5F5
+}
+
+' -- Core Auth --
+class "AuthConfig" as AC <<core>> {
+    +enabled: bool
+    +tokens: dict[str, str]
+    +check(token) → role | None
+    --
+    Static config-file tokens
+    hmac.compare_digest
+}
+
+class "AuthResult" as AR <<core>> {
+    +user_id: str
+    +scopes: frozenset[str]
+    +token_source: str
+    +has_scope(scope) → bool
+}
+
+class "check_request()" as CR <<core>> {
+    auth_config, method, path,
+    auth_header, cookie_header,
+    jwt_secret, storage
+    → (allowed, status, msg, AuthResult)
+    --
+    1. Auth disabled → allow
+    2. Public path → allow
+    3. Extract Bearer / cookie
+    4. Detect token type
+    5. Validate → AuthResult
+    6. Check scope vs path
+}
+
+' -- Token Types --
+class "JWT (HS256)" as JWT <<jwt>> {
+    sub: user_id
+    scopes: "read,write,approve"
+    src: "password" | "database"
+    iat, exp (24h default)
+    --
+    Detected by: contains "."
+    Validated locally
+    No DB call
+}
+
+class "API Token" as AT <<jwt>> {
+    Format: ts_ + 64 hex
+    Stored: SHA-256 hash
+    --
+    Detected by: starts with "ts_"
+    Lookup by hash in DB
+    Expiry check
+}
+
+class "Config Token" as CT <<core>> {
+    Raw value in memory
+    Role: "read" | "full"
+    --
+    Detected by: fallback
+    hmac.compare_digest
+    No DB needed
+}
+
+' -- Scopes --
+class "Scope Hierarchy" as SH <<scope>> {
+    read: {read}
+    write: {read, write}
+    approve: {read, write, approve}
+    --
+    GET → read
+    POST write paths → write
+    POST /api/approve → approve
+    /api/admin/* → approve
+}
+
+' -- Storage --
+class "users" as UT <<storage>> {
+    user_id (PK)
+    username (unique)
+    display_name
+    password_hash (bcrypt)
+    created
+}
+
+class "api_tokens" as TT <<storage>> {
+    token_id (PK)
+    token_hash (SHA-256, unique)
+    token_prefix
+    user_id → users
+    name, scopes
+    created, expires
+}
+
+' -- Endpoints --
+class "POST /api/auth/login" as Login <<endpoint>> {
+    {username, password}
+    OR {token: "ts_xxx"}
+    → {jwt, role, scopes, user_id}
+    --
+    Sets HttpOnly cookie
+}
+
+class "GET /api/auth/status" as Status <<endpoint>> {
+    → {auth_enabled, has_users,
+       setup_required}
+    --
+    Public (no auth)
+    Drives UI setup wizard
+}
+
+class "POST /api/auth/setup" as Setup <<endpoint>> {
+    {username, display_name, password}
+    → {jwt, user_id, scopes}
+    --
+    Public (no auth)
+    Only when zero users exist
+    Returns 409 if already set up
+}
+
+class "Admin API (Console)" as Admin <<endpoint>> {
+    POST/GET/DELETE users
+    POST/GET tokens
+    DELETE tokens/{id}
+    --
+    Requires approve scope
+}
+
+' -- Relationships --
+CR --> AC : config tokens
+CR --> JWT : validate
+CR --> AT : hash lookup
+CR --> CT : hmac check
+CR --> AR : returns
+CR --> SH : checks
+
+Login --> JWT : issues
+Login --> UT : verify password
+Login --> TT : verify API token
+
+Setup --> UT : create first user
+Setup --> JWT : issues
+
+AT --> TT : lookup by hash
+Admin --> UT : CRUD
+Admin --> TT : CRUD
+
+AR --> SH : scopes from
+
+JWT ..> AR : produces
+AT ..> AR : produces
+CT ..> AR : produces
+
+note right of CR
+  **Middleware Flow**
+  AuthMiddleware on every request:
+  1. Extract token from header/cookie
+  2. Detect type (JWT / ts_ / config)
+  3. Validate → AuthResult
+  4. Set ctx_user_id for logging
+  5. Store auth_result in scope state
+end note
+
+note bottom of SH
+  **Console** owns admin endpoints
+  **Server** validates JWT + config only
+  Both share JWT signing secret
+end note
+
+@enduml

docs/diagrams/16-channel-architecture.puml

@@ -0,0 +1,250 @@
+@startuml
+!theme plain
+title Turnstone — Channel Integration Architecture
+
+skinparam class {
+    BackgroundColor<<platform>> #E1BEE7
+    BackgroundColor<<service>> #E8EAF6
+    BackgroundColor<<mq>> #FFCDD2
+    BackgroundColor<<bridge>> #C8E6C9
+    BackgroundColor<<server>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+}
+
+' -- External Platforms --
+class "Discord" as Discord <<platform>> {
+    Gateway WebSocket (v10)
+    Message events
+    Interaction callbacks (buttons)
+    Thread-per-workstream
+    --
+    discord.py 2.x
+    asyncio event loop
+}
+
+class "Slack (future)" as Slack <<platform>> {
+    Socket Mode / Events API
+    Block Kit messages
+    --
+    Planned integration
+}
+
+class "Teams (future)" as Teams <<platform>> {
+    Bot Framework
+    Adaptive Cards
+    --
+    Planned integration
+}
+
+' -- Channel Service --
+class "turnstone-channel" as ChannelService <<service>> {
+    entry point: turnstone-channel
+    --
+    One process per platform
+    asyncio event loop
+    Structured logging (structlog)
+    --log-level, --log-format
+    --
+    POST /v1/api/notify (HTTP)
+    GET  /health
+}
+
+class "DiscordBot" as Bot <<service>> {
+    +on_message(msg)
+    +on_interaction(interaction)
+    +send(channel_id, content)
+    +run(token)
+    --
+    discord.py Client
+    Receives message events
+    Sends replies + embeds
+    Creates threads for workstreams
+    Renders approval buttons
+    escape_mentions() on send
+}
+
+class "ChannelRouter" as Router <<service>> {
+    +resolve_route(platform, channel_id)
+        → ws_id | None
+    +register_route(channel_id, ws_id)
+    +resolve_identity(platform, platform_user_id)
+        → user_id | None
+    --
+    Maps channels → workstreams
+    Maps platform users → turnstone users
+    Caches routes in memory
+}
+
+class "AsyncRedisBroker" as Broker <<service>> {
+    +push_inbound(msg)
+    +subscribe(ws_id) → AsyncIterator
+    +subscribe_global() → AsyncIterator
+    +push_response(correlation_id, msg)
+    --
+    redis.asyncio client
+    Pub/sub + queue operations
+}
+
+' -- Redis MQ --
+class "Redis MQ" as Redis <<mq>> {
+    turnstone:inbound (LIST)
+    turnstone:events:{ws_id} (PUBSUB)
+    turnstone:events:global (PUBSUB)
+    turnstone:resp:{corr_id} (LIST)
+    --
+    Shared message bus
+    Same queues as bridge protocol
+}
+
+' -- Bridge + Server --
+class "turnstone-bridge" as Bridge <<bridge>> {
+    BLPOP turnstone:inbound
+    Drive server via HTTP
+    Relay SSE → Redis pub/sub
+    --
+    Owns workstream lifecycle
+    Auto-approve / manual approve
+}
+
+class "turnstone-server" as Server <<server>> {
+    POST /v1/api/send
+    POST /v1/api/approve
+    POST /v1/api/workstreams/new
+    GET  /v1/api/events?ws_id=
+    --
+    LLM execution + tool use
+    SSE event stream
+    --
+    notify tool: _exec_notify()
+    ServiceTokenManager (JWT)
+}
+
+' -- Storage --
+class "channel_users" as CU <<storage>> {
+    channel_user_id (PK)
+    platform: "discord" | "slack"
+    platform_user_id
+    user_id → users
+    linked_at
+    --
+    /link command creates row
+    Resolved on each inbound message
+}
+
+class "channel_routes" as CR <<storage>> {
+    channel_type (PK)
+    channel_id (PK)
+    ws_id
+    node_id
+    created
+    --
+    Maps platform channels
+    to turnstone workstreams
+}
+
+class "services" as SVC <<storage>> {
+    service_type (PK)
+    service_id (PK)
+    url
+    last_heartbeat
+    created
+    --
+    Heartbeat every 30s
+    Stale after 120s
+    ON CONFLICT DO UPDATE
+}
+
+' -- Relationships --
+Discord --> Bot : gateway\nevents
+Bot --> Router : on_message\non_interaction
+Router --> Broker : SendMessage\nApproveMessage
+Router --> CU : resolve identity
+Router --> CR : resolve / register route
+Broker --> Redis : RPUSH inbound\nRPUSH resp:{id}
+
+Redis --> Bridge : BLPOP inbound
+Bridge --> Server : HTTP API
+Server --> Bridge : SSE events
+Bridge --> Redis : PUBLISH events:{ws_id}\nPUBLISH events:global
+
+Redis --> Broker : SUBSCRIBE events:{ws_id}
+Broker --> Bot : event stream
+Bot --> Discord : reply / embed\nbutton callback
+
+Slack .[hidden]. Discord
+Teams .[hidden]. Slack
+
+ChannelService --> Bot : creates + runs
+ChannelService --> Router : creates
+ChannelService --> Broker : creates
+ChannelService --> SVC : register / heartbeat /\nderegister
+
+' -- Notification path (direct HTTP, bypasses MQ) --
+Server --> ChannelService : POST /v1/api/notify\n(JWT: aud=turnstone-channel)
+Server --> SVC : list_services("channel",\nmax_age_seconds=120)
+
+' -- Notes --
+note right of Bot
+  **Inbound Flow**
+  1. Discord message arrives via gateway
+  2. Bot.on_message() fires
+  3. ChannelRouter resolves channel → ws_id
+     (or creates new workstream)
+  4. ChannelRouter resolves platform user → user_id
+     via channel_users table
+  5. Broker.push_inbound(SendMessage)
+  6. Bridge pops from Redis, drives server
+
+  **Workstream Resume (evicted workstreams)**
+  1. Stale route detected (no MQ owner)
+  2. Existing ws_id reused directly from route
+  3. CreateWorkstreamMessage sent with
+     resume_ws=<ws_id>
+  4. Server resumes atomically during creation
+  5. Bridge emits WorkstreamResumedEvent → thread
+end note
+
+note right of Broker
+  **Outbound Flow**
+  1. Server emits SSE events
+  2. Bridge relays to Redis events:{ws_id}
+  3. Broker.subscribe(ws_id) yields events
+  4. Bot formats and sends to Discord thread
+end note
+
+note bottom of CR
+  **Approval Flow**
+  1. ApprovalRequestEvent arrives via events:{ws_id}
+  2. Bot renders Discord buttons (Approve / Deny)
+  3. User clicks button → on_interaction()
+  4. Router builds ApproveMessage
+  5. Broker.push_response(correlation_id, msg)
+  6. Bridge pops from resp:{id}, calls POST /api/approve
+end note
+
+note bottom of CU
+  **Identity Linking**
+  1. User runs /link in Discord
+  2. Bot opens modal requesting API token
+  3. User submits ts_... API token
+  4. Bot validates token against storage
+  5. On success, inserts channel_users row
+  6. Subsequent messages carry user_id
+  7. AuthResult scopes applied by server
+end note
+
+note bottom of SVC
+  **Notification Flow** (direct HTTP, bypasses MQ)
+  1. LLM calls notify tool → _prepare_notify()
+  2. _exec_notify() checks rate limit (5/turn)
+  3. Queries services table for healthy gateways
+  4. Mints JWT (aud: turnstone-channel) via
+     ServiceTokenManager
+  5. POSTs to first healthy gateway
+  6. Gateway validates JWT, resolves target
+  7. adapter.send() → Discord API
+  8. On failure: retry up to 3× (1s, 3s backoff)
+  9. SSRF: only http(s) URLs allowed
+end note
+
+@enduml

docs/diagrams/17-notify-flow.puml

@@ -0,0 +1,116 @@
+@startuml
+!theme plain
+title Turnstone — Notification Delivery Flow
+
+skinparam participant {
+    BackgroundColor<<server>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<service>> #E8EAF6
+    BackgroundColor<<platform>> #E1BEE7
+}
+
+participant "ChatSession\n(turnstone-server)" as Session <<server>>
+participant "StorageBackend" as Storage <<storage>>
+participant "ServiceTokenManager" as STM <<server>>
+participant "Channel Gateway\n(_http.py)" as Gateway <<service>>
+participant "ChannelAdapter\n(Discord bot)" as Adapter <<service>>
+participant "Discord API" as Discord <<platform>>
+
+== Prepare Phase ==
+
+Session -> Session : _prepare_notify(call_id, args)
+note right
+  Validates:
+  - message (required, ≤2000 chars)
+  - target: username OR channel_type+channel_id
+  - no ambiguous targeting (both set)
+  - partial targeting errors
+end note
+
+== Execute Phase ==
+
+Session -> Session : _exec_notify(item)
+Session -> Session : check rate limit\n(≥5 per turn?)
+
+alt rate limit exceeded
+    Session --> Session : "Error: rate limit exceeded"
+end
+
+loop up to 3 attempts (retry delays: 1s, 3s)
+
+    Session -> Storage : list_services("channel",\nmax_age_seconds=120)
+    Storage --> Session : services[] (sorted by\nlast_heartbeat DESC)
+
+    alt no healthy services
+        Session -> Session : log.warning("notify.no_services")
+        Session -> Session : sleep(delay)
+    else services available
+
+        Session -> STM : bearer_header
+        note right
+          Lazy-init ServiceTokenManager
+          aud: turnstone-channel
+          scope: write
+          Auto-rotates 1h JWTs
+        end note
+        STM --> Session : Authorization: Bearer <jwt>
+
+        loop for each gateway (first-healthy)
+            Session -> Session : SSRF check:\nurl.startswith("http://"|"https://")
+
+            Session -> Gateway : POST /v1/api/notify\n+ Authorization header
+            Gateway -> Gateway : _check_auth()\nvalidate JWT (aud=turnstone-channel)\nor static token
+
+            alt auth failed
+                Gateway --> Session : 401 Unauthorized
+            else auth ok
+
+                alt username target
+                    Gateway -> Storage : get_user_by_username()
+                    Storage --> Gateway : user
+                    Gateway -> Storage : list_channel_users_by_user()
+                    Storage --> Gateway : linked channels
+                else direct target
+                    Gateway -> Gateway : use channel_type + channel_id
+                end
+
+                Gateway -> Adapter : send(channel_id, content)
+                note right
+                  escape_mentions() applied
+                  Chunked for 2000-char limit
+                end note
+                Adapter -> Discord : POST message
+                Discord --> Adapter : message_id
+                Adapter --> Gateway : message_id
+                Gateway --> Session : 200 {results: [{status: "sent"}]}
+
+                Session -> Session : _notify_count += 1
+                Session --> Session : "Notification sent successfully"
+                note right : Return — no further\ngateways tried
+            end
+        end
+
+        alt all gateways failed
+            Session -> Session : log.warning(\n"notify.all_gateways_failed")
+            Session -> Session : sleep(delay)
+        end
+
+    end
+end
+
+alt all retries exhausted
+    Session -> Session : log.warning("notify.delivery_failed")
+    Session --> Session : "Error: notification delivery failed"
+end
+
+== Service Registry (Background) ==
+
+note over Gateway, Storage
+  **Heartbeat Lifecycle**
+  1. Gateway startup: register_service("channel", id, url)
+  2. Every 30s: heartbeat_service("channel", id)
+  3. Shutdown: deregister_service("channel", id)
+  4. Stale after 120s (4 missed heartbeats)
+end note
+
+@enduml

docs/diagrams/18-watch-architecture.puml

@@ -0,0 +1,166 @@
+@startuml
+!theme plain
+title Turnstone — Watch Tool Architecture
+
+skinparam participant {
+    BackgroundColor<<server>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<ui>> #E8EAF6
+}
+
+participant "ChatSession\n(session.py)" as Session <<session>>
+participant "WatchRunner\n(watch.py)" as Runner <<server>>
+participant "StorageBackend\n(SQLite)" as Storage <<storage>>
+participant "WebUI / SSE\n(server.py)" as UI <<ui>>
+
+== Create Phase ==
+
+Session -> Session : _prepare_watch(action="create")
+note right
+  Validates:
+  - command via is_command_blocked()
+  - poll_every → parse_duration()
+  - stop_on → validate_condition()
+  - max watches limit (5)
+  - duplicate name check
+  needs_approval = True
+end note
+
+Session -> Storage : create_watch(watch_id, ws_id,\nnode_id, command, interval,\nstop_on, max_polls, next_poll)
+
+Session --> UI : tool_result:\n"Watch 'pr-review' created"
+
+== Poll Phase (WatchRunner daemon, every 15s) ==
+
+Runner -> Storage : list_due_watches(now)
+Storage --> Runner : due_watches[]
+note right
+  Filters:
+  active=1 AND
+  next_poll <= now AND
+  node_id matches
+end note
+
+loop for each due watch
+
+    Runner -> Runner : is_command_blocked()?
+    alt blocked
+        Runner -> Storage : update_watch(active=False)
+    else safe
+
+        Runner -> Runner : subprocess.run(command)
+        note right
+          timeout = tool_timeout
+          start_new_session = True
+          output truncated at 64KB
+        end note
+
+        Runner -> Runner : evaluate_condition(\nstop_on, output,\nexit_code, prev_output)
+        note right
+          **Variables:**
+          output, data, exit_code,
+          prev_output, changed
+
+          **Safe builtins only:**
+          len, str, int, sorted, ...
+          No import/open/exec/eval
+
+          **stop_on=None:**
+          fires on change (skip 1st poll)
+        end note
+
+        alt condition fired OR max_polls reached
+            Runner -> Storage : update_watch(\npoll_count++,\nlast_output, active=False)
+            Runner -> Runner : format_watch_message()
+            Runner -> Runner : _dispatch_result(ws_id, msg)
+        else not fired
+            Runner -> Storage : update_watch(\npoll_count++,\nlast_output, next_poll)
+        end
+
+    end
+end
+
+== Dispatch Phase ==
+
+note over Runner, Session
+  **Three dispatch paths:**
+end note
+
+alt Path A: workstream active + idle
+    Runner -> Session : dispatch_fn(message)\n→ _watch_pending.put()
+    Session -> Session : _dispatch_pending_watch()\n→ self.send(message)
+    Session -> UI : SSE: thinking, content,\ntool calls...
+    note right
+      Watch result appears as
+      synthetic user message.
+      Model sees it and responds.
+      Depth guard: max 5 chains.
+    end note
+
+else Path B: workstream active + busy
+    Runner -> Session : dispatch_fn(message)\n→ _watch_pending.put()
+    note right
+      Queued. Dispatched when
+      current send() reaches IDLE.
+    end note
+
+else Path C: workstream evicted
+    Runner -> Runner : restore_fn(ws_id)
+    note right
+      1. mgr.create() — may evict
+         another idle workstream
+      2. session.resume(ws_id)
+      3. set_watch_runner()
+      4. register new dispatch_fn
+    end note
+    Runner -> Session : restored dispatch_fn(message)
+end
+
+== Cancel / List ==
+
+Session -> Storage : list_watches_for_ws(ws_id)
+note right : action="list" (auto-approve)
+
+Session -> Storage : update_watch(active=False)
+note right : action="cancel" (auto-approve)
+
+== Server Lifecycle ==
+
+note over Runner, Storage
+  **Startup:**
+  1. WatchRunner created in main() with storage + node_id
+  2. restore_fn closure captures WorkstreamManager
+  3. Initial workstream: session.set_watch_runner(runner)
+  4. _lifespan(): runner.start() — daemon thread begins
+
+  **New workstream:**
+  session.set_watch_runner(runner) in create_workstream()
+  → registers dispatch_fn for ws_id
+
+  **Eviction / close:**
+  session.close() → runner.remove_dispatch_fn(ws_id)
+  Watches remain active in DB — WatchRunner uses restore_fn
+
+  **Restart recovery:**
+  Overdue watches fire ONE immediate poll
+  next_poll updated to now + interval
+  Normal cadence resumes
+
+  **Shutdown:**
+  _lifespan(): runner.stop() — joins thread
+end note
+
+== REST API ==
+
+note over UI, Storage
+  **GET /v1/api/watches[?ws_id=X]**
+  List active watches (for node or workstream)
+
+  **POST /v1/api/watches/{watch_id}/cancel**
+  Cancel a watch (sets active=False)
+
+  Both require write scope
+end note
+
+@enduml

docs/diagrams/19-governance-architecture.puml

@@ -0,0 +1,98 @@
+@startuml
+!theme plain
+skinparam backgroundColor #FFFFFF
+skinparam defaultFontName "IBM Plex Mono"
+skinparam componentStyle rectangle
+
+title Turnstone Governance Architecture
+
+package "Auth Flow" {
+  [Login/Token Auth] as auth
+  [_load_user_permissions()] as perms
+  [_permissions_to_scopes()] as scopes
+  [create_jwt()] as jwt
+}
+
+package "Middleware" {
+  [AuthMiddleware\n(scope check)] as mw
+  [require_permission()\n(granular check)] as rp
+}
+
+package "Governance Storage" {
+  database "roles" as roles_db
+  database "user_roles" as ur_db
+  database "orgs" as orgs_db
+  database "tool_policies" as tp_db
+  database "prompt_templates" as pt_db
+  database "usage_events" as ue_db
+  database "audit_events" as ae_db
+  database "workstream_templates" as wt_db
+  database "workstream_template_versions" as wtv_db
+}
+
+package "Runtime Enforcement" {
+  [evaluate_tool_policies_batch()] as eval
+  [WebUI.approve_tools()] as approve
+  [record_usage_event()] as usage
+  [record_audit()] as audit
+}
+
+package "Template Runtime" {
+  [_load_templates()] as tload
+  [_render_template()\n{{model}}, {{ws_id}}, {{node_id}}] as trender
+  [_init_system_messages()] as tsys
+  [set_template() / /template] as tset
+}
+
+package "WS Template Runtime" {
+  [resolve_ws_template()] as wtr
+  [apply settings\n(model, budget, prompt)] as wta
+  [drift detection\n(prompt_template_hash)] as wtd
+  [budget gate\n(session.send)] as wtb
+}
+
+package "Console UI" {
+  [Admin Panel\n10 tabs] as ui
+  [governance.js] as govjs
+  [sessionStorage\npermissions] as ss
+}
+
+auth --> perms : user_id
+perms --> roles_db : JOIN user_roles + roles
+perms --> scopes : permission set
+scopes --> jwt : scopes + permissions
+
+jwt --> mw : JWT in cookie/header
+mw --> rp : scope OK → check permission
+
+rp --> ui : 403 or allow
+
+eval --> tp_db : list_tool_policies()
+approve --> eval : tool names
+approve --> ae_db : (via audit)
+
+usage --> ue_db : on_status()
+audit --> ae_db : admin handlers
+
+govjs --> roles_db : /v1/api/admin/roles
+govjs --> tp_db : /v1/api/admin/policies
+govjs --> pt_db : /v1/api/admin/templates
+govjs --> ue_db : /v1/api/admin/usage
+govjs --> ae_db : /v1/api/admin/audit
+
+tload --> pt_db : list_default_templates()\nor get_by_name()
+tload --> trender : template content
+trender --> tsys : rendered content
+tset --> tload : name or None
+
+govjs --> wt_db : /v1/api/admin/ws-templates
+wtr --> wt_db : get_ws_template_by_name()
+wtr --> wta : template settings
+wta --> pt_db : prompt_template lookup
+wtd --> wt_db : compare hash
+wtb --> approve : __budget_override__
+wtv_db <.. wt_db : version snapshots
+
+auth -[hidden]-> mw
+mw -[hidden]-> approve
+@enduml

docs/diagrams/20-mcp-architecture.puml

@@ -0,0 +1,158 @@
+@startuml
+!theme plain
+title Turnstone — MCP Architecture (Resources, Prompts, Tools)
+
+skinparam participant {
+    BackgroundColor<<mcp>> #E1BEE7
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<server>> #FFE0B2
+    BackgroundColor<<ui>> #E8EAF6
+}
+
+participant "MCP Server\n(external)" as MCPSrv <<mcp>>
+participant "MCPClientManager\n(mcp_client.py)" as MCPMgr <<mcp>>
+participant "ChatSession\n(session.py)" as Session <<session>>
+participant "StorageBackend\n(governance)" as Storage <<storage>>
+participant "Server / Console\n(health + UI)" as UI <<server>>
+
+== Startup: Connection & Discovery ==
+
+MCPMgr -> MCPSrv : initialize (stdio or HTTP)
+MCPSrv --> MCPMgr : capabilities\n(tools, resources, prompts)
+
+MCPMgr -> MCPSrv : tools/list
+MCPSrv --> MCPMgr : Tool[]
+
+opt resources capability
+  MCPMgr -> MCPSrv : resources/list
+  MCPSrv --> MCPMgr : Resource[]
+  MCPMgr -> MCPSrv : resources/templates/list
+  MCPSrv --> MCPMgr : ResourceTemplate[]
+end
+
+opt prompts capability
+  MCPMgr -> MCPSrv : prompts/list
+  MCPSrv --> MCPMgr : Prompt[]
+end
+
+note over MCPMgr
+  Per-server storage:
+  _per_server_tools, _per_server_resources, _per_server_prompts
+  Copy-on-write rebuild into _tools, _resources, _prompts
+  Prefix: mcp__{server}__{name}
+end note
+
+MCPMgr -> Session : notify tool listeners
+MCPMgr -> Session : notify resource listeners
+
+== Governance Sync (on connect & refresh) ==
+
+MCPMgr -> Storage : sync_prompts_to_storage()
+note right
+  For each MCP prompt:
+  - Manual template exists? → skip
+  - MCP template exists? → update
+    (reset is_default=False)
+  - New? → create (origin="mcp",
+    readonly=True, is_default=False)
+  Removed prompts → delete
+  Protected by _sync_lock
+end note
+
+== set_storage() from entry point ==
+
+UI -> MCPMgr : set_storage(backend)
+note right
+  If servers already connected,
+  triggers immediate sync
+end note
+
+== Runtime: Tool Execution ==
+
+Session -> Session : _prepare_mcp_tool(func_name, args)
+note right
+  approval_label = func_name
+  (e.g. mcp__github__search)
+  needs_approval = True
+end note
+Session -> MCPMgr : call_tool_sync(name, args)
+MCPMgr -> MCPSrv : tools/call
+MCPSrv --> MCPMgr : ToolResult
+MCPMgr --> Session : output (text)
+
+== Runtime: Resource Read ==
+
+Session -> Session : _prepare_read_resource(uri)
+note right
+  approval_label = mcp_resource__{normalized_uri}
+  URI normalized (.. resolved)
+  needs_approval = True
+end note
+Session -> MCPMgr : read_resource_sync(uri)
+MCPMgr -> MCPSrv : resources/read
+MCPSrv --> MCPMgr : ReadResourceResult
+MCPMgr --> Session : content (text/blob)
+
+== Runtime: Prompt Invocation ==
+
+Session -> Session : _prepare_use_prompt(name, arguments)
+note right
+  approval_label = mcp__srv__prompt
+  Validated via is_mcp_prompt()
+  needs_approval = True
+end note
+Session -> MCPMgr : get_prompt_sync(name, args)
+MCPMgr -> MCPSrv : prompts/get
+MCPSrv --> MCPMgr : GetPromptResult
+MCPMgr --> Session : messages [{role, content}]
+
+== Three-Tier Refresh ==
+
+group Push Notifications
+  MCPSrv -> MCPMgr : ToolListChangedNotification
+  MCPMgr -> MCPMgr : _refresh_server_tools()
+
+  MCPSrv -> MCPMgr : ResourceListChangedNotification
+  MCPMgr -> MCPMgr : _refresh_server_resources()
+
+  MCPSrv -> MCPMgr : PromptListChangedNotification
+  MCPMgr -> MCPMgr : _refresh_server_prompts()
+  MCPMgr -> Storage : sync_prompts_to_storage()
+end
+
+group Periodic Polling (default 4h)
+  MCPMgr -> MCPMgr : _periodic_refresh()
+  note right
+    Only polls capabilities
+    without push support.
+    Staggered per-server.
+  end note
+end
+
+group Manual Refresh
+  Session -> MCPMgr : refresh_sync()
+  note right: /mcp refresh [server]
+end
+
+== Policy Evaluation ==
+
+note over Session
+  Tool policies use fnmatch on approval_label:
+  - mcp__github__* → allow (all GitHub tools/prompts)
+  - mcp_resource__file:///docs/* → allow
+  - mcp_resource__* → deny (block all resource reads)
+  - mcp__untrusted__* → ask
+end note
+
+== UI Visibility ==
+
+UI -> MCPMgr : server_count, get_resources(), get_prompts()
+note over UI
+  /health → mcp.servers, mcp.resources, mcp.prompts
+  Server UI: magenta status badge
+  Console: cluster status bar + node detail
+  System message: <mcp-resources> + <mcp-prompts> catalogs
+end note
+
+@enduml

docs/diagrams/21-ws-template-architecture.puml

@@ -0,0 +1,161 @@
+@startuml
+!theme plain
+title Turnstone — Workstream Template Architecture
+
+skinparam participant {
+    BackgroundColor<<admin>> #E8EAF6
+    BackgroundColor<<server>> #FFE0B2
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<integration>> #F3E5F5
+}
+
+participant "Admin / Console UI\n(governance.js)" as Admin <<admin>>
+participant "Server\n(server.py)" as Server <<server>>
+participant "ChatSession\n(session.py)" as Session <<session>>
+participant "StorageBackend\n(SQLite)" as Storage <<storage>>
+participant "Integration Points\n(scheduler, channel,\nbridge, MQ)" as Integrations <<integration>>
+
+== Admin CRUD ==
+
+Admin -> Server : POST /v1/api/admin/ws-templates
+note right
+  **Payload:**
+  name, model, system_prompt,
+  temperature, reasoning_effort,
+  max_tokens, agent_max_turns,
+  auto_approve, auto_approve_tools,
+  token_budget, prompt_template,
+  prompt_template_hash, notify_on_complete
+end note
+
+Server -> Storage : create_ws_template()
+Storage --> Server : ws_template_id
+
+Admin -> Server : PUT /v1/api/admin/ws-templates/{id}
+Server -> Storage : get_ws_template(id)\n(snapshot pre-update state)
+Storage --> Server : existing template
+Server -> Storage : create_ws_template_version()\n(version snapshot)
+Server -> Storage : update_ws_template(id, ...)
+note right
+  **Versioning:**
+  Each update snapshots
+  pre-update state into
+  workstream_template_versions.
+  version counter increments.
+end note
+
+Admin -> Server : GET /v1/api/admin/ws-templates
+Server -> Storage : list_ws_templates()
+
+Admin -> Server : DELETE /v1/api/admin/ws-templates/{id}
+Server -> Storage : delete_ws_template(id)
+
+== Workstream Creation Flow ==
+
+Integrations -> Server : CreateWorkstreamMessage\n(ws_template="production-agent")
+note right
+  **Sources:**
+  - Console UI (Profile dropdown)
+  - Scheduler (ws_template field)
+  - Channel Router (ws_template)
+  - Bridge (ws_template forwarding)
+  - MQ Client (ws_template)
+end note
+
+Server -> Storage : get_ws_template_by_name("production-agent")
+Storage --> Server : template dict
+
+Server -> Server : resolve_ws_template()\napply model override
+note right
+  **Settings applied:**
+  - model (overrides default)
+  - system_prompt
+  - temperature
+  - reasoning_effort
+  - max_tokens
+  - agent_max_turns
+  - auto_approve / auto_approve_tools
+  - token_budget
+  - tool_search config
+end note
+
+Server -> Session : mgr.create(model=template.model, ...)
+Session -> Session : _init_system_messages()
+
+alt template has prompt_template
+    Session -> Storage : get_prompt_template_by_name()
+    Session -> Session : _render_template()\n{{model}}, {{ws_id}}, {{node_id}}
+end
+
+Session -> Storage : _save_config()\n+ ws_template_id, ws_template_version
+
+== Drift Detection ==
+
+Server -> Server : compute prompt_template_hash\n(at creation time)
+note right
+  **Hash stored:**
+  SHA-256 of prompt_template
+  content at ws creation time.
+  Compared at next creation
+  to detect upstream changes.
+end note
+
+Server -> Storage : update_workstream()\n(store prompt_template_hash)
+
+... later, new workstream created ...
+
+Server -> Storage : get_ws_template()
+Server -> Server : compare hash vs\ncurrent prompt_template content
+alt hash mismatch
+    Server -> Server : log.warning(\n"prompt template drift detected")
+end
+
+== Token Budget Enforcement ==
+
+Session -> Session : send(message)
+Session -> Session : _check_budget_gate()
+note right
+  **Budget gate:**
+  if token_budget set:
+    total = prompt_tokens + completion_tokens
+    if total >= token_budget:
+      block further sends
+end note
+
+alt budget exceeded
+    Session -> Session : approve_tools(\n__budget_override__)
+    note right
+      Model can request
+      budget override via
+      special approval label.
+      User must approve.
+    end note
+else within budget
+    Session -> Session : continue normal flow
+end
+
+== Storage Schema ==
+
+note over Storage
+  **workstream_templates**
+  id, name (unique), model, system_prompt,
+  temperature, reasoning_effort, max_tokens,
+  agent_max_turns, auto_approve, auto_approve_tools,
+  token_budget, prompt_template, prompt_template_hash,
+  tool_search, tool_search_threshold, tool_search_max_results,
+  version, created_at, updated_at
+
+  **workstream_template_versions**
+  id, template_id (FK), version, snapshot (JSON),
+  created_at
+
+  **workstreams** (updated columns)
+  + ws_template_id: str | None
+  + ws_template_version: int | None
+
+  **scheduled_tasks** (updated column)
+  + ws_template: str | None
+end note
+
+@enduml

docs/diagrams/22-judge-architecture.puml

@@ -0,0 +1,160 @@
+@startuml
+!theme plain
+title Turnstone — Intent Validation (Judge) Architecture
+
+skinparam participant {
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<judge>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<ui>> #E8EAF6
+    BackgroundColor<<fs>> #F5F5F5
+}
+
+participant "ChatSession\n(session.py)" as Session <<session>>
+participant "IntentJudge\n(judge.py)" as Judge <<judge>>
+participant "LLM Provider\n(provider)" as LLM <<judge>>
+participant "StorageBackend\n(SQLite)" as Storage <<storage>>
+participant "WebUI / SSE\n(server.py)" as UI <<ui>>
+participant "Filesystem" as FS <<fs>>
+
+== Tool Call Requires Approval ==
+
+Session -> Session : _prepare_tool_calls()
+note right
+  Tool calls parsed from
+  LLM response. Auto-approved
+  tools dispatched immediately.
+  Remaining items need approval.
+end note
+
+Session -> Session : _evaluate_intent(pending_items)
+
+== Tier 1: Heuristic (synchronous, sub-ms) ==
+
+Session -> Judge : evaluate(items, messages, callback)
+
+Judge -> Judge : evaluate_heuristic()\nfor each item
+note right
+  **Rule table (first match wins):**
+  Critical (0.90, deny): rm /, mkfs,
+    dd, pipe-to-shell, chmod 777 /,
+    write/edit /etc/ .ssh/
+  High (0.80, review): sudo, kill -9,
+    destructive git, DROP TABLE,
+    secrets, HTTP mutations, ssh/scp
+  Medium (0.70, review): pip/npm install,
+    write_file, MCP tools, docker ops
+  Low (0.85, approve): read_file,
+    list_directory, search, recall,
+    read-only bash (ls, cat, grep...)
+  Default: medium, 0.50, review
+end note
+
+Judge --> Session : heuristic_verdicts[]
+
+Session -> Session : attach _heuristic_verdict\nto each pending item
+
+Session -> UI : SSE: approve_request\n{items: [{verdict: ...}],\n judge_pending: true}
+note right
+  Heuristic verdict displayed
+  immediately as risk badge.
+  Spinner shown while LLM
+  judge evaluates.
+end note
+
+Session -> Storage : create_intent_verdict()\nfor each heuristic verdict
+
+== Tier 2: LLM Judge (daemon thread, async) ==
+
+Judge -> Judge : spawn daemon thread\n"intent-judge"
+
+note over Judge, LLM
+  **Context preparation:**
+  1. FIFO-truncate conversation history
+     to max_context_ratio of context window
+  2. Append tool call details as user message
+  3. System prompt defines judge role + JSON schema
+end note
+
+loop up to 3 turns (timeout budget)
+
+    Judge -> LLM : create_completion(\nmodel, judge_messages,\ntools=[read_file, list_directory])
+    LLM --> Judge : CompletionResult
+
+    alt tool_calls present (turn < 3)
+        Judge -> Judge : _exec_read_only_tool()
+        note right
+          **Security hardening:**
+          Blocked: /etc/, /root/,
+          /proc/, /sys/, /dev/,
+          .ssh, .gnupg, .aws,
+          *.pem, *.key, *.p12
+          File cap: 32KB
+          Dir cap: 200 entries
+        end note
+        Judge -> FS : read_file / list_directory
+        FS --> Judge : file contents
+        Judge -> Judge : append tool result\nto judge_messages
+    else text response (final verdict)
+        Judge -> Judge : _parse_verdict()
+        note right
+          **4-stage JSON parsing:**
+          1. Direct JSON.loads
+          2. Markdown code block
+          3. Brace-counting
+          4. Regex field extraction
+        end note
+    end
+
+end
+
+== Tier 3: Arbitration ==
+
+Judge -> Judge : compare confidence:\nLLM vs heuristic
+note right
+  Only deliver LLM verdict
+  if confidence > heuristic.
+  Otherwise heuristic stands.
+end note
+
+alt LLM confidence > heuristic confidence
+    Judge -> Session : callback(llm_verdict)
+    Session -> UI : SSE: intent_verdict\n{tier: "llm", ...}
+    note right
+      UI replaces heuristic badge
+      with LLM verdict. Spinner
+      resolves to final assessment.
+    end note
+    Session -> Storage : create_intent_verdict()\nfor LLM verdict
+end
+
+== User Decision ==
+
+UI -> Session : resolve_approval(\napproved, feedback)
+
+Session -> Storage : update_intent_verdict(\nverdict_id, user_decision)
+note right
+  All tracked verdicts
+  (heuristic + LLM) updated
+  with "approved" or "denied".
+  Swap-and-clear avoids racing
+  with daemon judge thread.
+end note
+
+== Lifecycle ==
+
+note over Session, Judge
+  **Lazy initialization:**
+  IntentJudge created on first approval if judge_config.enabled.
+  Re-uses session's provider/client by default (self-consistency).
+  Cross-model: separate provider/client from [judge] config.
+
+  **Sub-agent exemption:**
+  Plan agent and task agent skip intent validation entirely.
+
+  **Storage:**
+  intent_verdicts table (migration 012). Verdicts queryable via
+  GET /v1/api/admin/verdicts (requires admin.judge permission).
+end note
+
+@enduml

docs/diagrams/23-memory-architecture.puml

@@ -0,0 +1,159 @@
+@startuml
+!theme plain
+title Turnstone — Structured Memory Architecture
+
+skinparam participant {
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<facade>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<api>> #E8EAF6
+    BackgroundColor<<sdk>> #F5F5F5
+}
+
+participant "ChatSession\n(session.py)" as Session <<session>>
+participant "MemoryFacade\n(memory.py)" as Facade <<facade>>
+participant "MemoryRelevance\n(memory_relevance.py)" as Relevance <<facade>>
+participant "StorageBackend\n(SQLite)" as Storage <<storage>>
+participant "Server API\n(server.py)" as API <<api>>
+participant "Console Admin\n(console/server.py)" as Admin <<api>>
+participant "SDK Client\n(sdk/)" as SDK <<sdk>>
+
+== Phase 1: Tool Path (session.send) ==
+
+Session -> Session : _prepare_tool_calls()\nparse memory(action=...)
+note right
+  Tool schema: 4 actions
+  save, search, delete, list
+  Auto-approved (no approval needed)
+end note
+
+Session -> Session : _exec_memory(item)
+
+alt action = save
+    Session -> Facade : save_structured_memory(\nname, content, description,\nmem_type, scope, scope_id)
+    Facade -> Facade : normalize_key(name)
+    Facade -> Storage : create_structured_memory()
+    alt unique constraint violation
+        Storage --> Facade : IntegrityError
+        Facade -> Storage : get_structured_memory_by_name()
+        Storage --> Facade : existing row
+        Facade -> Storage : update_structured_memory()
+    end
+    Storage --> Facade : memory_id
+    Facade --> Session : (memory_id, old_content)
+    Session -> Session : _init_system_messages()\nrefresh BM25 context
+end
+
+alt action = search
+    Session -> Facade : search_structured_memories(\nquery, mem_type, scope,\nscope_id, limit)
+    Facade -> Storage : search_structured_memories()
+    Storage --> Session : matched rows
+end
+
+alt action = delete
+    Session -> Facade : delete_structured_memory(\nname, scope, scope_id)
+    Facade -> Storage : delete_structured_memory()
+    Storage --> Session : bool (existed)
+    Session -> Session : _init_system_messages()\nrefresh BM25 context
+end
+
+== Phase 2: BM25 Relevance Injection ==
+
+Session -> Session : _init_system_messages()\nevery conversation turn
+
+Session -> Session : _get_visible_memories(\nlimit=fetch_limit)
+note right
+  **Scope resolution:**
+  1. global scope (always)
+  2. workstream scope (ws_id)
+  3. user scope (user_id, if auth)
+  Combined and deduplicated.
+end note
+
+Session -> Facade : list_structured_memories()\nper scope
+Facade -> Storage : list_structured_memories()
+Storage --> Session : up to fetch_limit rows
+
+Session -> Relevance : extract_recent_context(\nmessages, max_messages=3)
+Relevance --> Session : user text context
+
+Session -> Relevance : score_memories(\nmemories, context,\nk=relevance_k)
+note right
+  **BM25 scoring:**
+  Index over name + description
+  + content[:200] for each memory.
+  Returns top-k by relevance.
+  Empty query returns most recent k.
+end note
+Relevance --> Session : top-k memories
+
+Session -> Relevance : build_memory_context(\nrelevant_memories)
+note right
+  Formats as XML block:
+  <memories>
+    <memory name="..." type="..."
+     scope="..." description="...">
+      content (max 500 chars)
+    </memory>
+  </memories>
+end note
+Relevance --> Session : XML string
+
+Session -> Session : inject into\nsystem message
+
+== Phase 3: Server API Path ==
+
+SDK -> API : GET /v1/api/memories\n?type=project&limit=20
+API -> Facade : list_structured_memories()
+Facade -> Storage : list_structured_memories()
+Storage --> API : rows
+API --> SDK : {"memories": [...], "total": N}
+
+SDK -> API : POST /v1/api/memories\n{name, content, ...}
+API -> API : validate type, scope,\nname length, content length
+API -> Facade : save_structured_memory()
+Facade -> Storage : create / update
+Storage --> API : memory row
+API --> SDK : 201 (created) / 200 (updated)
+
+SDK -> API : POST /v1/api/memories/search\n{query, type, ...}
+API -> Facade : search_structured_memories()
+Facade -> Storage : search_structured_memories()
+Storage --> API : matched rows
+API --> SDK : {"memories": [...], "total": N}
+
+SDK -> API : DELETE /v1/api/memories/{name}\n?scope=global
+API -> Facade : delete_structured_memory()
+Facade -> Storage : delete row
+API --> SDK : {"status": "ok"}
+
+== Phase 4: Console Admin Path ==
+
+SDK -> Admin : GET /v1/api/admin/memories\n?type=&scope=&limit=
+Admin -> Admin : require_permission(\n"admin.memories")
+Admin -> Storage : list_structured_memories()
+Storage --> Admin : rows
+Admin --> SDK : {"memories": [...], "total": N}
+
+SDK -> Admin : GET /v1/api/admin/memories/{id}
+Admin -> Storage : get_structured_memory(id)
+Storage --> Admin : memory row
+Admin --> SDK : memory JSON
+
+SDK -> Admin : DELETE /v1/api/admin/memories/{id}
+Admin -> Storage : delete_structured_memory_by_id()
+Admin -> Admin : record_audit(\n"memory.delete")
+Admin --> SDK : {"status": "ok"}
+
+== Configuration ==
+
+note over Session, Relevance
+  **MemoryConfig** (from [memory] in config.toml):
+  relevance_k = 5       -- top-k memories per turn
+  fetch_limit = 50       -- max memories fetched for scoring
+  max_content = 32768    -- max content length per memory
+  nudge_cooldown = 300   -- seconds between metacognitive nudges
+  nudges = true          -- enable/disable memory nudges
+end note
+
+@enduml

docs/diagrams/24-settings-architecture.puml

@@ -0,0 +1,151 @@
+@startuml
+!theme plain
+title Turnstone — Settings Architecture
+
+skinparam participant {
+    BackgroundColor<<session>> #C8E6C9
+    BackgroundColor<<config>> #FFE0B2
+    BackgroundColor<<storage>> #B3E5FC
+    BackgroundColor<<api>> #E8EAF6
+    BackgroundColor<<sdk>> #F5F5F5
+}
+
+participant "Server\n(main)" as Server <<session>>
+participant "ConfigStore\n(config_store.py)" as Store <<config>>
+participant "SettingsRegistry\n(settings_registry.py)" as Registry <<config>>
+participant "StorageBackend\n(SQLite)" as Storage <<storage>>
+participant "Console Admin\n(console/server.py)" as Admin <<api>>
+participant "SDK Client\n(sdk/)" as SDK <<sdk>>
+participant "ChatSession\n(session.py)" as Session <<session>>
+
+== Phase 1: Server Startup ==
+
+Server -> Server : parse_args()\nCLI flags override defaults
+Server -> Server : init_storage()\nSQLite / PostgreSQL
+
+Server -> Store ** : ConfigStore(storage, node_id)
+Store -> Storage : get_system_settings_bulk(node_id)
+note right
+  1. Load global settings (node_id="")
+  2. Overlay per-node settings
+  Returns {key: json_value} dict
+end note
+Storage --> Store : raw settings
+Store -> Registry : deserialize_value(key, json)\nper entry
+Registry --> Store : typed values
+Store -> Store : swap _cache atomically\nincrement _version
+
+Server -> Server : warn_migrated_settings()
+note right
+  Scans config.toml for keys
+  now managed by ConfigStore.
+  Logs warning for each overlap.
+end note
+
+Server -> Server : session_factory captures\nConfigStore reference
+
+== Phase 2: Settings Read (session creation) ==
+
+Server -> Session : session_factory(ws_id)
+Session -> Store : get("model.temperature")
+Store -> Store : cache[key] lookup\n(lock-free)
+alt key in cache
+    Store --> Session : stored value
+else key not in cache
+    Store -> Registry : SETTINGS[key].default
+    Registry --> Store : default value
+    Store --> Session : default value
+end
+note right of Session
+  Settings are captured once
+  at workstream creation.
+  Not re-read on every turn.
+end note
+
+== Phase 3: Admin API — List / Schema ==
+
+SDK -> Admin : GET /v1/api/admin/settings
+Admin -> Admin : require_permission(\n"admin.settings")
+Admin -> Store : all_effective()
+Store -> Store : merge cache with\nregistry defaults
+Store --> Admin : {key: effective_value}
+Admin -> Registry : SETTINGS (metadata)
+note right
+  Annotates each setting with:
+  type, default, description,
+  is_stored, is_secret, constraints,
+  changed_by, updated
+end note
+Admin --> SDK : {"settings": [...], "total": N}
+
+SDK -> Admin : GET /v1/api/admin/settings/schema
+Admin -> Admin : require_permission(\n"admin.settings")
+Admin -> Registry : SETTINGS catalog
+Admin --> SDK : {"settings": [...], "total": N}
+
+== Phase 4: Admin API — Update ==
+
+SDK -> Admin : PUT /v1/api/admin/settings/\nmodel.temperature\n{"value": 0.7}
+Admin -> Admin : require_permission(\n"admin.settings")
+Admin -> Registry : validate_key("model.temperature")
+Registry --> Admin : SettingDef
+alt is_secret == true
+    Admin --> SDK : 403 Forbidden
+else
+    Admin -> Registry : validate_value(key, 0.7)
+    note right
+      Type coercion: float(0.7)
+      Range check: 0.0 <= 0.7 <= 2.0
+      Choices check: (none for this key)
+    end note
+    Registry --> Admin : typed value
+    Admin -> Store : set(key, 0.7, changed_by="admin")
+    Store -> Registry : serialize_value(0.7)\n=> "0.7"
+    Store -> Storage : upsert_system_setting(\nkey, "0.7", node_id, ...)
+    Storage --> Store : ok
+    Store -> Store : swap _cache atomically
+    Admin -> Admin : record_audit(\n"setting.update")
+    Admin --> SDK : {"key": "...", "value": 0.7,\n"previous": 0.5}
+end
+
+== Phase 5: Admin API — Delete (reset to default) ==
+
+SDK -> Admin : DELETE /v1/api/admin/settings/\nmodel.temperature
+Admin -> Admin : require_permission(\n"admin.settings")
+Admin -> Store : delete("model.temperature")
+Store -> Registry : validate_key(key)
+Store -> Storage : delete_system_setting(key, node_id)
+Storage --> Store : bool (existed)
+Store -> Store : remove from cache,\nswap atomically
+Admin -> Admin : record_audit(\n"setting.delete")
+Admin --> SDK : {"status": "ok",\n"key": "...", "default": 0.5}
+
+== Phase 6: Hot Reload ==
+
+SDK -> Admin : POST /v1/api/_internal/\nconfig-reload
+Admin -> Store : reload()
+Store -> Storage : get_system_settings_bulk(node_id)
+Storage --> Store : all settings
+Store -> Store : rebuild cache,\nswap atomically,\nincrement _version
+note right
+  Existing sessions: unchanged
+  (frozen at creation time).
+  New sessions: pick up
+  updated values immediately.
+end note
+Admin --> SDK : {"status": "ok"}
+
+== Precedence Summary ==
+
+note over Server, Registry
+  **Server entry point:**
+  CLI flag  >  ConfigStore (database)  >  registry default
+
+  **CLI entry point:**
+  CLI flag  >  config.toml  >  argparse default
+
+  **Bootstrap settings** (database, Redis, auth, server bind):
+  Always from config.toml / env vars — never in ConfigStore.
+end note
+
+@enduml

docs/diagrams/architecture-overview.svg

@@ -0,0 +1,286 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 1200 540" font-family="-apple-system, BlinkMacSystemFont, 'Segoe UI', Helvetica, Arial, sans-serif">
+  <defs>
+    <!-- Arrowhead markers -->
+    <marker id="arrow" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#484f58"/>
+    </marker>
+    <marker id="arrow-blue" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#58a6ff"/>
+    </marker>
+    <marker id="arrow-green" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#3fb950"/>
+    </marker>
+    <marker id="arrow-orange" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#f0883e"/>
+    </marker>
+    <marker id="arrow-coral" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#f47067"/>
+    </marker>
+    <marker id="arrow-muted" viewBox="0 0 10 7" refX="10" refY="3.5" markerWidth="8" markerHeight="6" orient="auto-start-auto">
+      <path d="M 0 0 L 10 3.5 L 0 7 z" fill="#8b949e"/>
+    </marker>
+
+    <!-- Card shadow filter -->
+    <filter id="shadow" x="-4%" y="-4%" width="108%" height="112%">
+      <feDropShadow dx="0" dy="1" stdDeviation="2" flood-color="#000" flood-opacity="0.4"/>
+    </filter>
+  </defs>
+
+  <!-- Background -->
+  <rect width="1200" height="540" rx="8" fill="#0d1117"/>
+
+  <!-- Title -->
+  <text x="600" y="36" text-anchor="middle" fill="#e6edf3" font-size="15" font-weight="700" letter-spacing="3">TURNSTONE</text>
+  <text x="600" y="54" text-anchor="middle" fill="#8b949e" font-size="11" letter-spacing="1">SYSTEM ARCHITECTURE</text>
+
+  <!-- ==================== COLUMN HEADERS ==================== -->
+  <text x="90" y="86" text-anchor="middle" fill="#58a6ff" font-size="9" font-weight="600" letter-spacing="2">CLIENTS</text>
+  <text x="276" y="86" text-anchor="middle" fill="#3fb950" font-size="9" font-weight="600" letter-spacing="2">GATEWAYS</text>
+  <text x="480" y="86" text-anchor="middle" fill="#f0883e" font-size="9" font-weight="600" letter-spacing="2">MESSAGE QUEUE</text>
+  <text x="700" y="86" text-anchor="middle" fill="#f47067" font-size="9" font-weight="600" letter-spacing="2">CLUSTER NODES</text>
+  <text x="940" y="86" text-anchor="middle" fill="#f778ba" font-size="9" font-weight="600" letter-spacing="2">LLM PROVIDERS</text>
+
+  <!-- ==================== CLIENT BOXES ==================== -->
+  <!-- CLI -->
+  <g filter="url(#shadow)">
+    <rect x="30" y="108" width="120" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="30" y="108" width="120" height="5" rx="5" fill="#58a6ff"/>
+    <rect x="30" y="108" width="120" height="5" fill="#58a6ff"/>
+    <rect x="30" y="111" width="120" height="2" fill="#161b22"/>
+    <text x="90" y="130" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">CLI</text>
+    <text x="90" y="145" text-anchor="middle" fill="#8b949e" font-size="9">terminal REPL</text>
+  </g>
+
+  <!-- Browser UI -->
+  <g filter="url(#shadow)">
+    <rect x="30" y="174" width="120" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="30" y="174" width="120" height="5" rx="5" fill="#58a6ff"/>
+    <rect x="30" y="174" width="120" height="5" fill="#58a6ff"/>
+    <rect x="30" y="177" width="120" height="2" fill="#161b22"/>
+    <text x="90" y="196" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Browser UI</text>
+    <text x="90" y="211" text-anchor="middle" fill="#8b949e" font-size="9">HTTP + SSE</text>
+  </g>
+
+  <!-- SDK / API -->
+  <g filter="url(#shadow)">
+    <rect x="30" y="244" width="120" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="30" y="244" width="120" height="5" rx="5" fill="#58a6ff"/>
+    <rect x="30" y="244" width="120" height="5" fill="#58a6ff"/>
+    <rect x="30" y="247" width="120" height="2" fill="#161b22"/>
+    <text x="90" y="266" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">SDK / API</text>
+    <text x="90" y="281" text-anchor="middle" fill="#8b949e" font-size="9">programmatic</text>
+  </g>
+
+  <!-- Discord / Slack -->
+  <g filter="url(#shadow)">
+    <rect x="30" y="314" width="120" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="30" y="314" width="120" height="5" rx="5" fill="#58a6ff"/>
+    <rect x="30" y="314" width="120" height="5" fill="#58a6ff"/>
+    <rect x="30" y="317" width="120" height="2" fill="#161b22"/>
+    <text x="90" y="336" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Discord / Slack</text>
+    <text x="90" y="351" text-anchor="middle" fill="#8b949e" font-size="9">chat platforms</text>
+  </g>
+
+  <!-- ==================== GATEWAY BOXES ==================== -->
+  <!-- Console -->
+  <g filter="url(#shadow)">
+    <rect x="216" y="118" width="120" height="58" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="216" y="118" width="120" height="5" rx="5" fill="#3fb950"/>
+    <rect x="216" y="118" width="120" height="5" fill="#3fb950"/>
+    <rect x="216" y="121" width="120" height="2" fill="#161b22"/>
+    <text x="276" y="142" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Console</text>
+    <text x="276" y="157" text-anchor="middle" fill="#8b949e" font-size="9">dashboard + proxy</text>
+    <text x="276" y="169" text-anchor="middle" fill="#8b949e" font-size="9">cluster management</text>
+  </g>
+
+  <!-- Channel Gateway -->
+  <g filter="url(#shadow)">
+    <rect x="216" y="292" width="120" height="58" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="216" y="292" width="120" height="5" rx="5" fill="#3fb950"/>
+    <rect x="216" y="292" width="120" height="5" fill="#3fb950"/>
+    <rect x="216" y="295" width="120" height="2" fill="#161b22"/>
+    <text x="276" y="316" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Channel Gateway</text>
+    <text x="276" y="331" text-anchor="middle" fill="#8b949e" font-size="9">platform adapter</text>
+    <text x="276" y="343" text-anchor="middle" fill="#8b949e" font-size="9">Discord, Slack, ...</text>
+  </g>
+
+  <!-- ==================== REDIS MQ ==================== -->
+  <g filter="url(#shadow)">
+    <rect x="420" y="168" width="120" height="132" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="420" y="168" width="120" height="5" rx="5" fill="#f0883e"/>
+    <rect x="420" y="168" width="120" height="5" fill="#f0883e"/>
+    <rect x="420" y="171" width="120" height="2" fill="#161b22"/>
+    <text x="480" y="198" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Redis MQ</text>
+    <line x1="438" y1="210" x2="522" y2="210" stroke="#30363d" stroke-width="1"/>
+    <text x="480" y="228" text-anchor="middle" fill="#8b949e" font-size="9">inbound queues</text>
+    <text x="480" y="243" text-anchor="middle" fill="#8b949e" font-size="9">event pub/sub</text>
+    <text x="480" y="258" text-anchor="middle" fill="#8b949e" font-size="9">node heartbeats</text>
+    <text x="480" y="273" text-anchor="middle" fill="#8b949e" font-size="9">workstream routing</text>
+    <text x="480" y="288" text-anchor="middle" fill="#8b949e" font-size="9">cluster state</text>
+  </g>
+
+  <!-- ==================== CLUSTER NODES ==================== -->
+  <!-- Cluster outline -->
+  <rect x="598" y="100" width="204" height="310" rx="8" fill="none" stroke="#30363d" stroke-width="1" stroke-dasharray="4,3"/>
+  <text x="700" y="422" text-anchor="middle" fill="#30363d" font-size="9" letter-spacing="1">CLUSTER</text>
+
+  <!-- Node A -->
+  <g filter="url(#shadow)">
+    <rect x="614" y="118" width="170" height="100" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="614" y="118" width="170" height="5" rx="5" fill="#f47067"/>
+    <rect x="614" y="118" width="170" height="5" fill="#f47067"/>
+    <rect x="614" y="121" width="170" height="2" fill="#161b22"/>
+    <text x="699" y="142" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Node A</text>
+    <line x1="632" y1="152" x2="766" y2="152" stroke="#30363d" stroke-width="1"/>
+    <!-- Bridge -->
+    <rect x="626" y="162" width="70" height="28" rx="3" fill="#1c2128" stroke="#30363d" stroke-width="1"/>
+    <text x="661" y="180" text-anchor="middle" fill="#8b949e" font-size="9">bridge</text>
+    <!-- Server -->
+    <rect x="704" y="162" width="70" height="28" rx="3" fill="#1c2128" stroke="#30363d" stroke-width="1"/>
+    <text x="739" y="180" text-anchor="middle" fill="#8b949e" font-size="9">server</text>
+    <!-- Arrow bridge to server -->
+    <line x1="696" y1="176" x2="702" y2="176" stroke="#484f58" stroke-width="1" marker-end="url(#arrow)"/>
+    <!-- Tools label -->
+    <text x="699" y="206" text-anchor="middle" fill="#484f58" font-size="8">14 tools + MCP</text>
+  </g>
+
+  <!-- Node B -->
+  <g filter="url(#shadow)">
+    <rect x="614" y="238" width="170" height="100" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="614" y="238" width="170" height="5" rx="5" fill="#f47067"/>
+    <rect x="614" y="238" width="170" height="5" fill="#f47067"/>
+    <rect x="614" y="241" width="170" height="2" fill="#161b22"/>
+    <text x="699" y="262" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Node B</text>
+    <line x1="632" y1="272" x2="766" y2="272" stroke="#30363d" stroke-width="1"/>
+    <!-- Bridge -->
+    <rect x="626" y="282" width="70" height="28" rx="3" fill="#1c2128" stroke="#30363d" stroke-width="1"/>
+    <text x="661" y="300" text-anchor="middle" fill="#8b949e" font-size="9">bridge</text>
+    <!-- Server -->
+    <rect x="704" y="282" width="70" height="28" rx="3" fill="#1c2128" stroke="#30363d" stroke-width="1"/>
+    <text x="739" y="300" text-anchor="middle" fill="#8b949e" font-size="9">server</text>
+    <!-- Arrow bridge to server -->
+    <line x1="696" y1="296" x2="702" y2="296" stroke="#484f58" stroke-width="1" marker-end="url(#arrow)"/>
+    <!-- Tools label -->
+    <text x="699" y="326" text-anchor="middle" fill="#484f58" font-size="8">14 tools + MCP</text>
+  </g>
+
+  <!-- ==================== LLM PROVIDERS ==================== -->
+  <!-- OpenAI -->
+  <g filter="url(#shadow)">
+    <rect x="870" y="130" width="140" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="870" y="130" width="140" height="5" rx="5" fill="#f778ba"/>
+    <rect x="870" y="130" width="140" height="5" fill="#f778ba"/>
+    <rect x="870" y="133" width="140" height="2" fill="#161b22"/>
+    <text x="940" y="153" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">OpenAI</text>
+    <text x="940" y="167" text-anchor="middle" fill="#8b949e" font-size="9">GPT-5, o-series</text>
+  </g>
+
+  <!-- Anthropic -->
+  <g filter="url(#shadow)">
+    <rect x="870" y="196" width="140" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="870" y="196" width="140" height="5" rx="5" fill="#f778ba"/>
+    <rect x="870" y="196" width="140" height="5" fill="#f778ba"/>
+    <rect x="870" y="199" width="140" height="2" fill="#161b22"/>
+    <text x="940" y="219" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Anthropic</text>
+    <text x="940" y="233" text-anchor="middle" fill="#8b949e" font-size="9">Claude 4.5 / 4.6</text>
+  </g>
+
+  <!-- Local / vLLM -->
+  <g filter="url(#shadow)">
+    <rect x="870" y="262" width="140" height="46" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="870" y="262" width="140" height="5" rx="5" fill="#f778ba"/>
+    <rect x="870" y="262" width="140" height="5" fill="#f778ba"/>
+    <rect x="870" y="265" width="140" height="2" fill="#161b22"/>
+    <text x="940" y="285" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">Local / vLLM</text>
+    <text x="940" y="299" text-anchor="middle" fill="#8b949e" font-size="9">llama.cpp, NIM</text>
+  </g>
+
+  <!-- ==================== STORAGE ==================== -->
+  <g filter="url(#shadow)">
+    <rect x="614" y="450" width="170" height="52" rx="5" fill="#161b22" stroke="#30363d" stroke-width="1"/>
+    <rect x="614" y="450" width="170" height="5" rx="5" fill="#bc8cff"/>
+    <rect x="614" y="450" width="170" height="5" fill="#bc8cff"/>
+    <rect x="614" y="453" width="170" height="2" fill="#161b22"/>
+    <text x="699" y="476" text-anchor="middle" fill="#e6edf3" font-size="11" font-weight="600">PostgreSQL / SQLite</text>
+    <text x="699" y="492" text-anchor="middle" fill="#8b949e" font-size="9">conversations, memory, auth</text>
+  </g>
+  <text x="699" y="444" text-anchor="middle" fill="#bc8cff" font-size="9" font-weight="600" letter-spacing="2">STORAGE</text>
+
+  <!-- ==================== CONNECTION LINES ==================== -->
+
+  <!-- CLIENT -> GATEWAY connections -->
+  <!-- Browser -> Console -->
+  <line x1="150" y1="197" x2="214" y2="155" stroke="#58a6ff" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-blue)"/>
+  <!-- Discord -> Channel -->
+  <line x1="150" y1="337" x2="214" y2="325" stroke="#58a6ff" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-blue)"/>
+
+  <!-- CLI -> direct to Node A server (top path, curved) -->
+  <path d="M 150 131 C 200 131, 200 100, 400 100 L 400 100 C 500 100, 570 140, 612 168" stroke="#58a6ff" stroke-width="1.2" stroke-opacity="0.5" fill="none" stroke-dasharray="6,3" marker-end="url(#arrow-blue)"/>
+  <text x="370" y="96" fill="#484f58" font-size="8" text-anchor="middle">direct</text>
+
+  <!-- SDK -> Redis (direct push) -->
+  <line x1="150" y1="267" x2="418" y2="240" stroke="#58a6ff" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-blue)"/>
+
+  <!-- GATEWAY -> REDIS connections -->
+  <!-- Console -> Redis -->
+  <line x1="336" y1="160" x2="418" y2="200" stroke="#3fb950" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-green)"/>
+  <!-- Channel -> Redis -->
+  <line x1="336" y1="318" x2="418" y2="272" stroke="#3fb950" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-green)"/>
+
+  <!-- REDIS -> NODE connections -->
+  <!-- Redis -> Node A bridge -->
+  <line x1="540" y1="210" x2="624" y2="176" stroke="#f0883e" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-orange)"/>
+  <!-- Redis -> Node B bridge -->
+  <line x1="540" y1="260" x2="624" y2="296" stroke="#f0883e" stroke-width="1.2" stroke-opacity="0.6" marker-end="url(#arrow-orange)"/>
+
+  <!-- Console -> Node (proxy, dashed) -->
+  <path d="M 336 147 C 380 130, 500 108, 612 145" stroke="#3fb950" stroke-width="1" stroke-opacity="0.4" fill="none" stroke-dasharray="4,3" marker-end="url(#arrow-green)"/>
+  <text x="468" y="120" fill="#484f58" font-size="8" text-anchor="middle">proxy</text>
+
+  <!-- NODE -> LLM connections -->
+  <!-- Node A -> LLM providers -->
+  <line x1="784" y1="168" x2="868" y2="155" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.5" marker-end="url(#arrow-coral)"/>
+  <line x1="784" y1="176" x2="868" y2="219" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.3"/>
+  <line x1="784" y1="180" x2="868" y2="282" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.2"/>
+
+  <!-- Node B -> LLM providers -->
+  <line x1="784" y1="288" x2="868" y2="163" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.2"/>
+  <line x1="784" y1="296" x2="868" y2="222" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.3"/>
+  <line x1="784" y1="300" x2="868" y2="288" stroke="#f47067" stroke-width="1.2" stroke-opacity="0.5" marker-end="url(#arrow-coral)"/>
+
+  <!-- NODE -> STORAGE connections -->
+  <line x1="680" y1="338" x2="680" y2="448" stroke="#bc8cff" stroke-width="1.2" stroke-opacity="0.4" stroke-dasharray="4,3" marker-end="url(#arrow-muted)"/>
+  <line x1="718" y1="218" x2="718" y2="236" stroke="#484f58" stroke-width="1" stroke-opacity="0.3" stroke-dasharray="2,2"/>
+
+  <!-- Extensibility hint -->
+  <text x="699" y="392" text-anchor="middle" fill="#30363d" font-size="10">...</text>
+
+  <!-- Event flow: Bridges -> Redis (dashed, bidirectional feel) -->
+  <line x1="624" y1="186" x2="542" y2="220" stroke="#f0883e" stroke-width="1" stroke-opacity="0.3" stroke-dasharray="3,3"/>
+  <line x1="624" y1="286" x2="542" y2="250" stroke="#f0883e" stroke-width="1" stroke-opacity="0.3" stroke-dasharray="3,3"/>
+  <text x="574" y="242" fill="#484f58" font-size="7" text-anchor="middle">events</text>
+
+  <!-- ==================== FLOW LABELS ==================== -->
+  <!-- Interactive flow label -->
+  <rect x="30" y="395" width="10" height="10" rx="2" fill="none" stroke="#58a6ff" stroke-width="1.5" stroke-dasharray="3,2"/>
+  <text x="46" y="404" fill="#8b949e" font-size="9">interactive (direct)</text>
+
+  <!-- Queue flow label -->
+  <rect x="160" y="395" width="10" height="10" rx="2" fill="none" stroke="#f0883e" stroke-width="1.5"/>
+  <text x="176" y="404" fill="#8b949e" font-size="9">queue-driven</text>
+
+  <!-- Proxy/event label -->
+  <rect x="275" y="395" width="10" height="10" rx="2" fill="none" stroke="#3fb950" stroke-width="1.5" stroke-dasharray="3,2"/>
+  <text x="291" y="404" fill="#8b949e" font-size="9">proxy / events</text>
+
+  <!-- ==================== BOTTOM DETAILS ==================== -->
+  <line x1="30" y1="430" x2="1170" y2="430" stroke="#21262d" stroke-width="1"/>
+
+  <!-- Routing rules at bottom, left-aligned -->
+  <text x="44" y="456" fill="#30363d" font-size="9" font-weight="600" letter-spacing="1">ROUTING</text>
+  <circle cx="44" cy="474" r="3" fill="#f47067" opacity="0.6"/>
+  <text x="54" y="477" fill="#484f58" font-size="9">target_node set  &#x2192;  route to specific node queue</text>
+  <circle cx="44" cy="494" r="3" fill="#f0883e" opacity="0.6"/>
+  <text x="54" y="497" fill="#484f58" font-size="9">ws_id set  &#x2192;  route to owning node</text>
+  <circle cx="44" cy="514" r="3" fill="#58a6ff" opacity="0.6"/>
+  <text x="54" y="517" fill="#484f58" font-size="9">neither  &#x2192;  shared queue, any node picks up</text></svg>

docs/diagrams/png/01-system-context.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:341a8ab1483b1e0146878bd384a11d56bc78d29262de8262d06ef924317e2762
-size 139969
+oid sha256:d8ce6d2a43a991655c3f64a20b6e810fdb2f78eb767acc3d3d1b8d2c9f443181
+size 165011

docs/diagrams/png/02-package-structure.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29534422fc31eee613f70a479aa14de5278b98c49bb75fce7a63b72e248f1149
-size 323269
+oid sha256:c9daca81971ba7a8ed6736d23d5373c69435158fa6240b9880d14fc4759ab580
+size 329673

docs/diagrams/png/03-core-engine-classes.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f85f24081d32318e079d26a4855ebb6e66df349ca7c7c703db50788af528426b
-size 376282
+oid sha256:01fbb3338df6426cefc2811541a865f268673b4febf32f524c264d120bc068fa
+size 589546

docs/diagrams/png/04-conversation-turn.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a90dec1546dd0f8343e3e27cf6f235d95f3ffc375928d34d0df5d8f25d63e5ad
-size 269901
+oid sha256:da9d32000e3d92d92ce621661ced60f276f9b5be652f5ed6123b400505415f4a
+size 319702

docs/diagrams/png/05-tool-pipeline.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cb36b4924394cf54d6aaef454cced317b72e336e580cc6ac25cd4b9d0917bec5
-size 243422
+oid sha256:6dd3c923d1e1c49b5f91d8d342fb4b0d49a46d432460379ad146a9e3b075a05a
+size 277234

docs/diagrams/png/06-mq-protocol.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e660f453a3708d1a7d827f07cd967f122500eb1c4845f5a75cc1309091bce6af
-size 185796
+oid sha256:2229801220548e4794baa67e27a0a39dc7968c826a28fa8144763e678c8ed733
+size 192556

docs/diagrams/png/07-message-routing.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8cc7c94d5ac4862c3c09450346f0af923818e02ea0550cc8023f039fdc701179
-size 221528
+oid sha256:09535722ba975e47cf0557a40b6c481f125ff2022c396f79715c3bba9f715871
+size 222032

docs/diagrams/png/08-redis-key-schema.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d33b9b3affcdb07086b5aebca8a3b9c2b009cdfc6f360950a0e72e65fbcb8f17
-size 201602
+oid sha256:ed457b10b534b5fc2a5e190b281d7ded4dd1615da2229d67a373cf5dddccd059
+size 201601

docs/diagrams/png/09-workstream-states.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:adac93a0bb062d7199b819a600a0983ff011a75d16928fb80322cbb41f9284ea
-size 158866
+oid sha256:7896c6e041b6dbb89d034468fa980c8fe645df5eb969d45ef966ccc6399edac2
+size 200083

docs/diagrams/png/10-simulator-architecture.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69f201cff948cb0a19810b7c4ad26d346f869ee2dd3141eba4f353332efa2e21
-size 373649
+oid sha256:35cf3a6942f62dabcbbe012ac2f9e6f155332c894692981b076de5a25c1f3330
+size 374055

docs/diagrams/png/11-console-data-flow.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:793d7c2b28a751c6f467f2de788fcd462d3b8fd9cd5cb7adb5b32d78fb185394
-size 236004
+oid sha256:a74b4b8b5dbfb1a51a01100b731477968942b01218bad9451a3d5a9cb3003294
+size 411665

docs/diagrams/png/12-deployment.png

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0e6c1dfaef840d5228645aaad3637c973b2f71c372595814f3b743a991f5c6fc
-size 239128
+oid sha256:84524f4bc900708ac8adf081591d336f862830188eb8505e71a0f071b339d923
+size 252599

docs/diagrams/png/13-sdk-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7c3e40c10425d721f833390ae3531c09af501157fd3142531ba4eba86ff719d
+size 197112

docs/diagrams/png/14-storage-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb5e7c221f6b1ee1082b37da32e65c45b5e468014cf6881a210e5b4d8a8dca8b
+size 255736

docs/diagrams/png/15-auth-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af5ab3126bf685afe68e24bc4b0ed97371d0ebdb77bf4d76c0331ab120580cc0
+size 248809

docs/diagrams/png/16-channel-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1380065cbb5f95b5ea7dc6b2a00986c455b82888af60784980dffbd936460dcf
+size 431129

docs/diagrams/png/17-notify-flow.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0f6097840fccdbfe16cd5e4c9f5d063b2c36942460a944df68b8ec947e63ea3
+size 221452

docs/diagrams/png/18-watch-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96176a09e65e90dadc32d5e9ed778423842be89204d2cf382225f53a90cfaf01
+size 258547

docs/diagrams/png/19-governance-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4dac4948d928b4705936d73b4d159aa1e89315ec0397616ca914bbf19e7a1ce
+size 206479

docs/diagrams/png/20-mcp-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e6dc5142c7908314ce01229b3c4f13bf9450adcbb62a178838bd4cf81d9f4da
+size 250417

docs/diagrams/png/21-ws-template-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c06d7086d7965eb9fe333396f027133d42507cf120bfe8dc851c009a8768ec48
+size 284926

docs/diagrams/png/22-judge-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:feb31b9d05ea56544053ad00457c389acba977c07ecc08870960e6e0ca64aa11
+size 279971

docs/diagrams/png/23-memory-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c89628ed917dfd576c1af75c68fe5fed9beadaaee9dcea7aa7a1643867c4f1b9
+size 344323

docs/diagrams/png/24-settings-architecture.png

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83c0e6aad3eb19f6bc475a30a77215e801da3da5930f0462417fe7eb6eda6be2
+size 347144

docs/docker.md

@@ -27,6 +27,9 @@ Console dashboard: http://localhost:8090
 | `server` | 8080 | default | Web UI + chat workstreams + LLM |
 | `bridge` | — | default | Redis-to-HTTP bridge (multi-node routing) |
 | `console` | 8090 | default | Cluster dashboard |
+| `channel` | — | production | Channel gateway (Discord, Slack, etc.) |
+| `server-1`…`server-10` | — | cluster | 10-node server fleet (PostgreSQL required) |
+| `bridge-1`…`bridge-10` | — | cluster | Matching bridge fleet |
 | `sim` | — | sim | Multi-node cluster simulator |
 
 ## Profiles
@@ -37,6 +40,18 @@ Console dashboard: http://localhost:8090
 docker compose up
 ```
 
+**Production** — adds PostgreSQL and the channel gateway. Requires `POSTGRES_PASSWORD` and (for Discord) `TURNSTONE_DISCORD_TOKEN`:
+
+```bash
+docker compose --profile production up
+```
+
+**Cluster** — 10-node server/bridge fleet sharing PostgreSQL and Redis. Access all nodes via the console at `:8090`. Requires `POSTGRES_PASSWORD`:
+
+```bash
+docker compose --profile cluster up
+```
+
 **Sim** — adds the simulator. Can run alongside the full stack or standalone with just Redis and the console:
 
 ```bash
@@ -57,7 +72,7 @@ All configuration is via environment variables in `.env` (copy from `.env.exampl
 |----------|---------|-------------|
 | `LLM_BASE_URL` | `http://host.docker.internal:8000/v1` | OpenAI-compatible API URL |
 | `OPENAI_API_KEY` | `dummy` | API key (`dummy` for local servers) |
-| `TAVILY_API_KEY` | — | Web search API key (optional) |
+| `TAVILY_API_KEY` | — | Web search API key (only needed for local/vLLM models; Anthropic and OpenAI search models use native search) |
 
 ### Redis
 
@@ -84,8 +99,35 @@ All configuration is via environment variables in `.env` (copy from `.env.exampl
 
 | Variable | Default | Description |
 |----------|---------|-------------|
-| `TURNSTONE_AUTH_ENABLED` | — | Set to `1` to require Bearer token auth |
-| `TURNSTONE_AUTH_TOKEN` | — | Shared auth token for server/bridge/console |
+| `TURNSTONE_AUTH_ENABLED` | — | Set to `1` to require authentication |
+| `TURNSTONE_AUTH_TOKEN` | — | Config-file token for server/bridge/console (backward compat, works alongside JWT) |
+| `TURNSTONE_JWT_SECRET` | — | Secret key for signing JWTs (required when using user identity / JWT auth) |
+
+### Database
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `TURNSTONE_DB_BACKEND` | `sqlite` | Storage backend: `sqlite` or `postgresql` |
+| `TURNSTONE_DB_URL` | — | Database URL (e.g. `postgresql://user:pass@db:5432/turnstone`). For SQLite, defaults to `/data/.turnstone.db` |
+
+The database stores workstream history, user accounts, and API tokens. When using JWT auth, a database backend is required for user storage.
+
+> **First-time setup:** After deploying with auth enabled, create an initial admin user by running `turnstone-admin create-user` inside the container:
+>
+> ```bash
+> docker compose exec server turnstone-admin create-user --username admin --name "Admin"
+> ```
+>
+> You will be prompted to set a password. Use it to log in via the UI or SDK, then create additional users through the admin API. Pass `--token --scopes read,write,approve` to also generate an initial API token.
+
+### Channel Gateway
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `TURNSTONE_DISCORD_TOKEN` | — | Discord bot token (required to enable Discord adapter) |
+| `TURNSTONE_DISCORD_GUILD` | `0` | Restrict to a single Discord guild (0 = all guilds) |
+
+The channel service runs in the `production` profile. When `TURNSTONE_DISCORD_TOKEN` is set, the Discord adapter connects to the Discord Gateway and routes messages through Redis MQ to the bridge and server. See [Channel Integrations](channels.md) for full setup instructions including Discord application creation and user account linking.
 
 ### Simulator
 
@@ -101,13 +143,13 @@ All configuration is via environment variables in `.env` (copy from `.env.exampl
 
 ## Scaling
 
-Scale to multiple server/bridge pairs:
+For multi-node testing, use the `cluster` profile which provides 10 dedicated server+bridge pairs with unique node IDs (`node-1` through `node-10`), resource limits, and shared PostgreSQL:
 
 ```bash
-docker compose up --scale server=3 --scale bridge=3
+POSTGRES_PASSWORD=secret docker compose --profile cluster up
 ```
 
-Each bridge auto-generates a unique node ID from its container hostname. When scaling `server`, remove the host port mapping (or use a reverse proxy) to avoid port conflicts.
+The default `server` and `bridge` also run alongside the cluster nodes (11 total). All nodes are accessible via the console dashboard at `:8090`.
 
 ## Volumes
 
@@ -128,7 +170,7 @@ docker compose build
 docker compose build --no-cache
 ```
 
-All five entry points are installed in a single image: `turnstone-server`, `turnstone-bridge`, `turnstone-console`, `turnstone-sim`, `turnstone-eval`.
+All entry points are installed in a single image: `turnstone-server`, `turnstone-bridge`, `turnstone-console`, `turnstone-channel`, `turnstone-admin`, `turnstone-sim`, `turnstone-eval`.
 
 ## Cleanup
 

docs/governance.md

@@ -0,0 +1,214 @@
+# Governance
+
+Turnstone governance provides role-based access control (RBAC), tool execution
+policies, prompt templates, usage tracking, and audit logging for the admin
+console.
+
+## Architecture
+
+See [diagram: 19-governance-architecture.puml](diagrams/19-governance-architecture.puml).
+
+### RBAC (Roles & Permissions)
+
+The permission model has two layers:
+
+1. **Scopes** (legacy) — `read`, `write`, `approve`. Checked by `AuthMiddleware`
+   on every request based on URL path classification.
+2. **Permissions** (granular) — 15 permission strings checked per-endpoint by
+   `require_permission()`.
+
+**Built-in roles** (seeded by migration 008):
+
+| Role | Permissions |
+|------|-------------|
+| admin | read, write, approve, admin.users, admin.roles, admin.orgs, admin.policies, admin.templates, admin.audit, admin.usage, admin.schedules, admin.watches, tools.approve, workstreams.create, workstreams.close |
+| operator | read, write, workstreams.create, workstreams.close |
+| viewer | read |
+
+Custom roles can be created with any subset of the 15 valid permissions.
+
+**Auth flow:**
+1. User logs in (password or API token) → `_load_user_permissions()` aggregates
+   permissions from all assigned roles
+2. `_permissions_to_scopes()` derives legacy scopes (any `admin.*` → `approve`)
+3. JWT created with both `scopes` and `permissions` claims
+4. Middleware checks scope → handler checks permission via `require_permission()`
+
+### Tool Policies
+
+Admin-defined rules that control tool execution:
+
+- **Pattern matching**: Glob syntax via `fnmatch` (e.g., `bash*`, `file_write`, `*`)
+- **Actions**: `allow` (auto-approve), `deny` (block), `ask` (normal approval flow)
+- **Priority**: Higher priority evaluated first, first match wins
+- **Enforcement**: `evaluate_tool_policies_batch()` called in `WebUI.approve_tools()`
+  before the `auto_approve` check
+- **MCP granular policies**: MCP resources and prompts are evaluated using their
+  `approval_label` for fine-grained control:
+  - Resource reads: `mcp_resource__{uri}` (e.g., `mcp_resource__file:///docs/*` to allow,
+    `mcp_resource__*` to deny all)
+  - Prompt invocations: `mcp__{server}__{prompt}` (e.g., `mcp__trusted__*` to allow,
+    `mcp__*` to require approval for all)
+  - Built-in tools continue to use `func_name` for backward compatibility
+
+### Prompt Templates
+
+Admin-curated system message templates injected at workstream startup:
+
+- **Runtime behavior**: Templates are loaded once at session creation and injected
+  into the system message *before* user `instructions`. Templates set the baseline;
+  instructions customize per-workstream behavior.
+- **Default templates**: All `is_default=true` templates auto-apply to new
+  workstreams, concatenated in alphabetical order by name. Use name prefixes
+  (e.g. `01-safety`, `02-style`) to control ordering.
+- **Explicit selection**: `--template <name>` CLI flag, `template` field on
+  `POST /v1/api/workstreams/new`, console creation modal dropdown, scheduled task
+  config, and channel adapter config. An explicit template *replaces* defaults.
+- **Variables**: Three built-in placeholders resolved at load time:
+  `{{model}}` (active model name), `{{ws_id}}` (workstream ID),
+  `{{node_id}}` (server node ID). Unrecognized placeholders are kept as-is.
+- **Runtime switching**: `/template <name>` to switch, `/template clear` to revert
+  to defaults, `/template` to show current. Persisted across resume.
+- **Categories**: general, engineering, support, custom, mcp
+- **Content limit**: 32 KB per template (enforced on create/update)
+- **Storage**: `prompt_templates` table with JSON `variables` array. Migration 010
+  adds `template` column to `scheduled_tasks`.
+- **MCP sync**: MCP server prompts auto-sync into prompt_templates with
+  `origin="mcp"`, `mcp_server` set, and `readonly=True`. Manual templates take
+  precedence on name collision. MCP-synced content updates reset `is_default` to
+  prevent compromised servers from injecting defaults. Admin UI shows origin badge
+  and disables edit/delete for MCP-sourced templates.
+
+### Workstream Templates
+
+Workstream templates are behavioral profiles applied at workstream creation — the next level beyond prompt templates. While prompt templates inject system message text, workstream templates define the complete workstream configuration.
+
+**What they define:**
+- System prompt (inline text OR reference to a prompt template by name)
+- Model override (empty = server default)
+- Temperature, reasoning effort, max tokens, agent max turns
+- Auto-approve policy (blanket and/or per-tool list)
+- Token budget (0 = unlimited; warns at 80%, requires approval at 100%)
+- Completion notification config (stored for v2 dispatch)
+
+**Storage:** `workstream_templates` table (migration 011) with auto-versioning. Edits snapshot the pre-update state into `workstream_template_versions`. Workstreams record which template and version spawned them via `ws_template_id` + `ws_template_version` columns.
+
+**Applied once at creation:** Template settings are snapshot-applied to the workstream's config. Not a live binding — template updates don't affect running workstreams.
+
+**Prompt template drift detection:** When a workstream template references a prompt template, a SHA-256 hash of the prompt content is stored at ws_template create/update time. At workstream creation, the server compares the stored hash against current content and logs a warning on mismatch.
+
+**Admin API:** 7 endpoints under `/v1/api/admin/ws-templates` (list, create, get, update, delete, version history) plus a read-only summary at `/v1/api/ws-templates`. Permission: `admin.ws_templates`.
+
+**Console UI:** "WS Templates" tab with CRUD table, create/edit modals (name, description, system prompt source toggle, model, auto-approve, per-tool auto-approve, temperature, reasoning effort, max tokens, agent max turns, token budget, enabled), and version history modal. "Profile" dropdown on workstream creation modal. "WS Template" dropdown on scheduler create/edit modals.
+
+**Token budget enforcement:** Tracked in `session.send()`. At 80% consumption, emits an info message. At 100%, the next turn requires explicit approval via the `__budget_override__` synthetic tool name (reuses existing approval UI — inline in browser, Discord buttons, bridge auto-approve). The synthetic name can be targeted by tool policies (e.g. `__budget_override__` → `allow` for admins).
+
+**SDK:** Python (`list_ws_templates`, `create_ws_template`, `get_ws_template`, `update_ws_template`, `delete_ws_template`, `list_ws_template_versions`) and TypeScript (`listWsTemplates`, `createWsTemplate`, etc.) on both sync and async console clients. `ws_template` parameter on `create_workstream()` for both server and console SDKs.
+
+### Usage Tracking
+
+Per-LLM-request token and tool call metrics:
+
+- **Recording**: `on_status()` in `WebUI` records a `usage_event` after each
+  LLM response with prompt/completion tokens, tool call count, model, ws_id
+- **Querying**: `GET /v1/api/admin/usage` with `group_by` (day/hour/model/user)
+  and time range filtering
+- **Pruning**: `prune_usage_events(retention_days=90)` and
+  `prune_audit_events(retention_days=365)` run automatically via the
+  console scheduler's periodic cleanup cycle
+
+### Audit Logging
+
+Append-only trail of admin actions:
+
+- **Recording**: `record_audit()` helper called from all admin mutation handlers
+- **Events captured**: user.create, user.delete, token.create, token.revoke,
+  channel.link, channel.unlink, role.create, role.update, role.delete,
+  role.assign, role.unassign, policy.create, policy.update, policy.delete,
+  template.create, template.update, template.delete,
+  ws_template.create, ws_template.update, ws_template.delete, org.update
+- **Querying**: `GET /v1/api/admin/audit` with action/user/time filters + pagination
+
+## Database Schema
+
+Migration 008 adds 7 tables:
+
+| Table | Purpose |
+|-------|---------|
+| `orgs` | Organizations (single default org for now) |
+| `roles` | Named permission bundles (3 builtin + custom) |
+| `user_roles` | User-to-role assignments (composite PK) |
+| `tool_policies` | Per-tool approve/deny/ask rules |
+| `prompt_templates` | Reusable system message templates |
+| `usage_events` | Per-request token/tool metrics |
+| `audit_events` | Admin action log |
+
+Also adds `org_id` column to `users` table.
+
+## API Endpoints
+
+All under `/v1/api/admin/` (requires `approve` scope + granular permission).
+
+| Group | Endpoints | Permission |
+|-------|-----------|------------|
+| Users / Tokens / Channels | 9 (CRUD) | `admin.users` |
+| Roles | 7 (CRUD + assignment) | `admin.roles` / `admin.users` |
+| Orgs | 3 (list, get, update) | `admin.orgs` |
+| Tool Policies | 4 (CRUD) | `admin.policies` |
+| Prompt Templates | 4 (CRUD) | `admin.templates` |
+| Schedules | 6 (CRUD + runs) | `admin.schedules` |
+| WS Templates | 7 (CRUD + versions + summary) | `admin.ws_templates` |
+| Watches | 3 (list, create, cancel) | `admin.watches` |
+| Usage | 1 (aggregated query) | `admin.usage` |
+| Audit | 1 (paginated, filtered) | `admin.audit` |
+
+Full OpenAPI spec at `/openapi.json` and Swagger UI at `/docs`.
+
+## Admin Console UI
+
+6 new tabs added to the admin panel (11 total):
+
+- **Roles** — CRUD roles, permission checkbox grid, user role assignment modal
+- **Policies** — CRUD tool policies with colored action badges (green/red/amber)
+- **Templates** — CRUD prompt templates with wide modal, textarea editor
+- **WS Templates** — CRUD workstream templates with create/edit modals, version history
+- **Usage** — Summary readouts + CSS bar chart, time range + group-by selectors
+- **Audit** — Filterable log with relative timestamps, load-more pagination
+
+Tabs are permission-gated: hidden if the user lacks the required permission.
+
+## SDK
+
+Both Python and TypeScript console SDKs expose governance methods:
+
+**Python** (`TurnstoneConsole` / `AsyncTurnstoneConsole`):
+- `list_roles()`, `create_role()`, `update_role()`, `delete_role()`
+- `list_user_roles()`, `assign_role()`, `unassign_role()`
+- `list_orgs()`, `get_org()`, `update_org()`
+- `list_policies()`, `create_policy()`, `update_policy()`, `delete_policy()`
+- `list_templates()`, `create_template()`, `update_template()`, `delete_template()`
+- `list_ws_templates()`, `create_ws_template()`, `get_ws_template()`, `update_ws_template()`, `delete_ws_template()`, `list_ws_template_versions()`
+- `get_usage(since, group_by=...)`, `get_audit(action=..., limit=...)`
+
+**TypeScript** (`TurnstoneConsole`):
+- Same methods with camelCase naming and typed interfaces
+
+## Security Considerations
+
+- **Privilege escalation prevented**: `admin_assign_role` blocks self-assignment
+  and requires caller to hold a superset of the target role's permissions
+- **Permission validation**: Role create/update validates permissions against
+  a 15-item allowlist (`_VALID_PERMISSIONS`)
+- **Self-deletion blocked**: `admin_delete_user` rejects attempts to delete
+  your own account (matching the self-assignment guard on role endpoints)
+- **Field allowlists**: Storage `update_*` methods filter fields against
+  allowlists (`_ROLE_MUTABLE`, `_POLICY_MUTABLE`, etc.) — handler bugs
+  cannot overwrite `role_id`, `builtin`, `created`, or other protected columns
+- **Bootstrap safety**: `handle_auth_setup` fails and rolls back if admin role
+  assignment fails, preventing locked-out first user
+- **API token RBAC**: `_authenticate_api_token` loads permissions from user's
+  roles, ensuring API tokens are subject to RBAC enforcement
+- **Policy evaluation is fail-open**: If storage is unavailable, tool policies
+  degrade to the existing approval flow (not auto-approve)
+- **Audit IP resolution**: `_audit_context()` prefers `X-Forwarded-For` for
+  client IP when behind a reverse proxy, falling back to `request.client.host`

docs/judge.md

@@ -0,0 +1,279 @@
+# Intent Validation (Judge)
+
+> See also: [Judge Architecture diagram](diagrams/png/22-judge-architecture.png)
+
+Intent validation provides advisory risk assessments for tool calls that require
+human approval. An LLM judge evaluates each tool call and presents a structured
+verdict alongside the approval prompt, helping users make informed decisions.
+
+## Overview
+
+When a tool call requires approval, the intent validation system runs a two-tier
+evaluation:
+
+1. **Heuristic tier** (instant) -- Pattern-based risk classification using a
+   rule table. Zero cost, sub-millisecond latency.
+2. **LLM judge tier** (async) -- Semantic evaluation using an LLM with
+   read-only tool access. Runs on a daemon thread and delivers its verdict
+   progressively.
+
+The verdict is purely advisory -- the user always makes the final decision.
+
+The heuristic verdict is attached to the `approve_request` SSE event immediately.
+The LLM verdict arrives later via an `intent_verdict` SSE event, allowing the
+UI to show a spinner that resolves into a richer assessment. Both verdicts are
+persisted to the `intent_verdicts` table for audit and future calibration.
+
+---
+
+## Configuration
+
+### config.toml
+
+```toml
+[judge]
+enabled = true
+model = ""                    # empty = same as session model
+provider = ""                 # empty = same as session provider
+base_url = ""
+api_key = ""
+confidence_threshold = 0.7   # reserved for v2 smart approvals (not used in v1)
+max_context_ratio = 0.5       # max % of judge context window for history
+timeout = 60.0                # seconds (generous for local models)
+read_only_tools = true        # judge can use read_file/list_directory
+```
+
+All fields are optional. The judge is enabled by default; use `enabled = false`
+(or `--no-judge` on the command line) to disable it.
+
+### CLI flags
+
+```
+--judge / --no-judge           Enable/disable (default: enabled)
+--judge-model MODEL            Model for judge
+--judge-provider PROVIDER      Provider for judge
+--judge-timeout SECONDS        LLM judge timeout (default: 60)
+--judge-confidence FLOAT       Confidence threshold (default: 0.7)
+```
+
+CLI flags override `config.toml` values.
+
+---
+
+## Judge Model Selection
+
+- **Default (self-consistency)**: When `model` is empty, the session model
+  evaluates its own tool calls. Research shows self-consistency achieves
+  comparable accuracy to multi-agent debate at a fraction of the cost.
+- **Cross-model**: Use a different model for the judge (e.g. local model for
+  the session, commercial model for the judge). Set `model` and `provider`
+  in the `[judge]` config section, or use `--judge-model` / `--judge-provider`
+  CLI flags.
+- **Cross-provider**: When both `model` and `provider` are set, the judge
+  creates its own LLM client. You can optionally specify `base_url` and
+  `api_key` for non-default endpoints.
+
+---
+
+## Heuristic Rules
+
+The heuristic tier scans a priority-ordered rule table (critical first, low
+last) and returns the first matching rule. Each rule has:
+
+- **Tool pattern**: fnmatch glob matched against `func_name` and `approval_label`
+- **Argument patterns**: Regex patterns matched against the tool's primary
+  argument text (command string for bash, path for file tools, JSON for others)
+- **Risk level, confidence, and recommendation**: Pre-assigned per rule
+
+### Rule tiers
+
+| Tier     | Confidence | Recommendation | Examples |
+|----------|-----------|----------------|----------|
+| Critical | 0.90      | deny           | `rm -rf /`, `mkfs`, `dd if=`, pipe-to-shell, chmod 777 on root, write/edit to `/etc/`, `.ssh/` |
+| High     | 0.80      | review         | `sudo`, `kill -9`, destructive git (`reset --hard`, `push --force`, `clean -f`), DROP TABLE, write/edit secrets (`.env`, `.pem`, `.key`), HTTP mutations, `ssh`/`scp` |
+| Medium   | 0.70      | review         | Package installs (`pip`, `npm`, `apt`, `brew`, `cargo`), `write_file` (default), MCP tool calls, Docker operations |
+| Low      | 0.85      | approve        | `read_file`, `list_directory`, `search`, `recall`, `man`, `use_prompt`, read-only bash commands (`ls`, `cat`, `head`, `grep`, `find`, etc.) |
+
+When no rule matches, the heuristic returns a default verdict: medium risk,
+0.50 confidence, "review" recommendation.
+
+The bash "read-only" rule handles simple pipelines and command chains by
+splitting on `|`, `&&`, `||`, and `;`, then checking each segment individually.
+
+---
+
+## LLM Judge
+
+The LLM judge runs on a daemon thread and performs a multi-turn evaluation:
+
+1. **Context preparation**: Recent conversation history is FIFO-truncated to
+   fit within `max_context_ratio` of the judge's context window. The tool call
+   details (name, approval label, full arguments) are appended as a user message.
+2. **Multi-turn loop** (up to 5 turns): The judge can use `read_file` and
+   `list_directory` to gather evidence before rendering its verdict. Each tool
+   result is appended to the conversation and the judge is called again. On
+   the final turn, tools are stripped and a forcing message instructs the
+   judge to render its verdict immediately.
+3. **Verdict parsing**: The judge's final text response is parsed as JSON using
+   a four-stage strategy: direct parse, markdown code block extraction,
+   brace-counting, and regex field extraction as a last resort.
+4. **Arbitration**: If the LLM verdict has higher confidence than the heuristic,
+   it replaces the heuristic via the `intent_verdict` SSE event.
+
+### Read-only tools
+
+When `read_only_tools` is enabled (default), the judge can use two tools:
+
+- **`read_file`**: Read file contents (capped at 32 KB)
+- **`list_directory`**: List directory entries (capped at 200 entries)
+
+Security hardening blocks access to sensitive paths:
+
+| Category | Blocked patterns |
+|----------|-----------------|
+| System directories | `/etc/`, `/root/`, `/proc/`, `/sys/`, `/dev/` |
+| Credential directories | `.ssh`, `.gnupg`, `.aws`, `.config` |
+| Key files | `*.pem`, `*.key`, `*.p12`, `*.pfx` |
+
+### Timeout
+
+The `timeout` setting (default 60 seconds) is a total budget across all judge
+turns. Time is decremented after each LLM call. If the budget expires mid-turn,
+the judge attempts to parse whatever partial response is available.
+
+---
+
+## Verdict Structure
+
+Each verdict (heuristic or LLM) is an `IntentVerdict` with these fields:
+
+| Field            | Type       | Description |
+|------------------|------------|-------------|
+| `verdict_id`     | string     | Unique identifier (UUID prefix) |
+| `call_id`        | string     | Correlates with the tool call's `call_id` |
+| `func_name`      | string     | Tool function name |
+| `intent_summary` | string     | One-sentence description of what the tool call does |
+| `risk_level`     | string     | `"low"`, `"medium"`, `"high"`, or `"critical"` |
+| `confidence`     | float      | 0.0--1.0, how certain the assessment is |
+| `recommendation` | string     | `"approve"`, `"review"`, or `"deny"` |
+| `reasoning`      | string     | Explanation of the assessment |
+| `evidence`       | list[str]  | Supporting evidence (rule name or file excerpts) |
+| `tier`           | string     | `"heuristic"` or `"llm"` |
+| `judge_model`    | string     | Model used (empty for heuristic tier) |
+| `latency_ms`     | int        | Evaluation time in milliseconds |
+
+---
+
+## Session Integration
+
+The judge is lazy-initialized on first use. When `ChatSession` prepares tool
+calls for approval, it calls `_evaluate_intent()` which:
+
+1. Instantiates `IntentJudge` if not already created
+2. Extracts `func_name`, `func_args`, and `approval_label` from each pending item
+3. Calls `judge.evaluate()` which returns heuristic verdicts immediately
+4. Attaches each heuristic verdict to its item as `_heuristic_verdict`
+5. The daemon thread runs the LLM judge and delivers results via `ui.on_intent_verdict()`
+
+Sub-agents (plan agent, task agent) are exempt from intent validation -- they
+always get full tool visibility without judge evaluation.
+
+---
+
+## Storage and Audit
+
+All verdicts are persisted to the `intent_verdicts` table (migration 012):
+
+- Heuristic verdicts are stored when the `approve_request` event is emitted
+- LLM verdicts are stored when the `intent_verdict` event is delivered
+- The `user_decision` column is updated when the user approves or denies
+
+The console admin panel exposes verdict history via:
+
+```
+GET /v1/api/admin/verdicts?ws_id=&since=&until=&risk_level=&limit=100&offset=0
+```
+
+This endpoint requires the `admin.judge` permission.
+
+---
+
+## SSE Events
+
+### `approve_request` (extended)
+
+When the judge is active, `approve_request` items include a `verdict` field
+with the heuristic verdict, and the event includes a `judge_pending` flag
+indicating that an LLM verdict is in flight:
+
+```json
+{
+  "type": "approve_request",
+  "judge_pending": true,
+  "items": [
+    {
+      "call_id": "call_abc123",
+      "header": "bash: npm install express",
+      "preview": "",
+      "func_name": "bash",
+      "approval_label": "bash",
+      "needs_approval": true,
+      "error": null,
+      "verdict": {
+        "verdict_id": "a1b2c3d4e5f6",
+        "call_id": "call_abc123",
+        "func_name": "bash",
+        "intent_summary": "Package installation: npm install express",
+        "risk_level": "medium",
+        "confidence": 0.70,
+        "recommendation": "review",
+        "reasoning": "Command installs a software package which may modify the environment.",
+        "evidence": ["Matched rule: package-install"],
+        "tier": "heuristic",
+        "judge_model": "",
+        "latency_ms": 0
+      }
+    }
+  ]
+}
+```
+
+### `intent_verdict`
+
+Delivered asynchronously when the LLM judge completes. The UI replaces the
+heuristic verdict badge with the LLM verdict:
+
+```json
+{
+  "type": "intent_verdict",
+  "verdict_id": "f7e8d9c0b1a2",
+  "call_id": "call_abc123",
+  "func_name": "bash",
+  "intent_summary": "Install Express.js web framework via npm",
+  "risk_level": "medium",
+  "confidence": 0.85,
+  "recommendation": "review",
+  "reasoning": "The command installs express from npm. This is a well-known package but will modify node_modules and package.json.",
+  "evidence": ["Checked package.json — express is not currently a dependency"],
+  "tier": "llm",
+  "judge_model": "gpt-5",
+  "latency_ms": 2340
+}
+```
+
+---
+
+## v2 Calibration Path
+
+Run v1 with all tools requiring manual approval to build a local verdict
+dataset. The `intent_verdicts` table accumulates `(tool_call, verdict,
+user_decision)` triples over time. In v2, calibration tooling will analyze
+this dataset to:
+
+- Identify tools that are always approved (candidates for auto-approve policies)
+- Detect false positives in heuristic rules
+- Measure LLM judge accuracy against human decisions
+- Recommend policy changes to reduce approval fatigue
+
+This data-driven approach means v1 is both useful on its own and a foundation
+for automated policy tuning.

docs/memory.md

@@ -0,0 +1,569 @@
+# Structured Memory
+
+> See also: [Memory Architecture diagram](diagrams/png/23-memory-architecture.png)
+
+The structured memory system gives the AI persistent, typed, scoped memories
+that survive across sessions and workstreams. Memories are automatically
+surfaced in the system message via BM25 relevance scoring, so the model has
+contextual recall without explicit search.
+
+## Overview
+
+Each memory has three dimensions:
+
+- **Type** -- categorizes the memory's purpose
+- **Scope** -- controls visibility boundaries
+- **Name** -- unique identifier within a scope (snake_case, normalized)
+
+### Memory types
+
+| Type        | Purpose                                                    |
+|-------------|------------------------------------------------------------|
+| `user`      | User preferences, conventions, working style               |
+| `project`   | Project-specific knowledge, architecture, patterns         |
+| `feedback`  | Corrections, lessons learned, things to avoid              |
+| `reference` | Reference material, documentation, specifications         |
+
+### Memory scopes
+
+| Scope        | Visibility                                                |
+|--------------|-----------------------------------------------------------|
+| `global`     | Visible to all workstreams and users                      |
+| `workstream` | Visible only within the originating workstream            |
+| `user`       | Follows the authenticated user across workstreams         |
+
+A memory's identity is the tuple `(name, scope, scope_id)`. Saving a memory
+with the same identity upserts -- updating content while preserving the ID.
+
+### BM25 relevance injection
+
+On every conversation turn, the system:
+
+1. Fetches up to `fetch_limit` memories visible in the current scope
+2. Extracts context from the last 3 user messages
+3. Scores memories against that context using a BM25 index
+4. Injects the top `relevance_k` memories into the system message as
+   `<memories>` XML tags
+5. Appends a hint telling the model how many memories are in scope
+
+This means the model always has its most relevant memories available without
+explicit recall -- but can still use `memory(action='search')` for deeper
+lookup.
+
+### Nudges
+
+The metacognition layer can nudge the model to save memories at appropriate
+moments (e.g., after a correction or when resuming a workstream). Nudges are
+rate-limited by `nudge_cooldown` and can be disabled entirely.
+
+---
+
+## Configuration
+
+### config.toml
+
+```toml
+[memory]
+relevance_k = 5          # top-k memories injected per turn
+fetch_limit = 50          # max memories fetched from storage for scoring
+max_content = 32768       # max content length per memory (characters)
+nudge_cooldown = 300      # minimum seconds between memory nudges
+nudges = true             # enable/disable metacognitive nudges
+```
+
+All fields are optional. Defaults are shown above.
+
+---
+
+## Tool Usage
+
+The `memory` tool supports four actions:
+
+### save
+
+Store or update a memory.
+
+```json
+{
+  "action": "save",
+  "name": "project_architecture",
+  "content": "The project uses a hexagonal architecture with...",
+  "description": "Core architecture patterns",
+  "type": "project",
+  "scope": "global"
+}
+```
+
+| Parameter     | Required | Default     | Description                              |
+|---------------|----------|-------------|------------------------------------------|
+| `name`        | yes      | --          | Snake_case identifier (max 256 chars)    |
+| `content`     | yes      | --          | Memory content (max `max_content` chars) |
+| `description` | no       | `""`        | Short description for relevance matching |
+| `type`        | no       | `"project"` | One of: user, project, feedback, reference |
+| `scope`       | no       | `"global"`  | One of: global, workstream, user         |
+
+### search
+
+Find memories by query (BM25 full-text search).
+
+```json
+{
+  "action": "search",
+  "query": "authentication patterns",
+  "type": "project",
+  "limit": 10
+}
+```
+
+| Parameter | Required | Default | Description                          |
+|-----------|----------|---------|--------------------------------------|
+| `query`   | yes      | --      | Search query                         |
+| `type`    | no       | `""`    | Filter by type                       |
+| `scope`   | no       | `""`    | Filter by scope                      |
+| `limit`   | no       | `20`    | Max results (capped at 50)           |
+
+### delete
+
+Remove a memory by name.
+
+```json
+{
+  "action": "delete",
+  "name": "outdated_pattern",
+  "scope": "global"
+}
+```
+
+| Parameter  | Required | Default    | Description              |
+|------------|----------|------------|--------------------------|
+| `name`     | yes      | --         | Memory name to delete    |
+| `scope`    | no       | `"global"` | Scope of the memory      |
+
+### list
+
+List all memories with optional filters.
+
+```json
+{
+  "action": "list",
+  "type": "feedback",
+  "limit": 50
+}
+```
+
+| Parameter | Required | Default | Description                |
+|-----------|----------|---------|----------------------------|
+| `type`    | no       | `""`    | Filter by type             |
+| `scope`   | no       | `""`    | Filter by scope            |
+| `limit`   | no       | `20`    | Max results (capped at 50) |
+
+---
+
+## Server API
+
+Four endpoints on the server for programmatic memory access.
+
+### `GET /v1/api/memories`
+
+List memories with optional filters.
+
+**Query parameters:**
+
+| Parameter  | Type   | Required | Default | Description                  |
+|------------|--------|----------|---------|------------------------------|
+| `type`     | string | no       | `""`    | Filter by memory type        |
+| `scope`    | string | no       | `""`    | Filter by scope              |
+| `scope_id` | string | no       | `""`    | Filter by scope ID           |
+| `limit`    | int    | no       | `100`   | Max results (capped at 200)  |
+
+When `scope=user` and `scope_id` is omitted, the authenticated user's ID is
+used automatically.
+
+**Response:** `200`
+
+```json
+{
+  "memories": [
+    {
+      "memory_id": "a1b2c3d4-e5f6-...",
+      "name": "project_architecture",
+      "description": "Core architecture patterns",
+      "type": "project",
+      "scope": "global",
+      "scope_id": "",
+      "content": "The project uses a hexagonal architecture...",
+      "created": "2026-03-10T10:00:00",
+      "updated": "2026-03-12T14:30:00"
+    }
+  ],
+  "total": 1
+}
+```
+
+---
+
+### `POST /v1/api/memories`
+
+Save or upsert a structured memory.
+
+**Request body:**
+
+```json
+{
+  "name": "deployment_process",
+  "content": "Deploy via GitHub Actions. Staging auto-deploys on push to main.",
+  "description": "CI/CD deployment workflow",
+  "type": "project",
+  "scope": "global",
+  "scope_id": ""
+}
+```
+
+| Field        | Type   | Required | Default     | Description                          |
+|--------------|--------|----------|-------------|--------------------------------------|
+| `name`       | string | yes      | --          | Memory name (max 256 chars)          |
+| `content`    | string | yes      | --          | Memory content (max 65536 chars)     |
+| `description`| string | no       | `""`        | Short description for search ranking |
+| `type`       | string | no       | `"project"` | One of: user, project, feedback, reference |
+| `scope`      | string | no       | `"global"`  | One of: global, workstream, user     |
+| `scope_id`   | string | no       | `""`        | Scope qualifier (auto-resolved for user scope) |
+
+**Response (created):** `201`
+
+```json
+{
+  "memory_id": "a1b2c3d4-e5f6-...",
+  "name": "deployment_process",
+  "description": "CI/CD deployment workflow",
+  "type": "project",
+  "scope": "global",
+  "scope_id": "",
+  "content": "Deploy via GitHub Actions...",
+  "created": "2026-03-14T10:00:00",
+  "updated": "2026-03-14T10:00:00"
+}
+```
+
+**Response (updated):** `200` -- same schema, returned when a memory with the
+same `(name, scope, scope_id)` already existed.
+
+**Errors:**
+
+| Status | Condition                          |
+|--------|------------------------------------|
+| 400    | Missing name, empty content, invalid type/scope, content too long |
+
+---
+
+### `POST /v1/api/memories/search`
+
+Search memories by query. Uses POST for the request body but is non-mutating
+(requires only `read` scope).
+
+**Request body:**
+
+```json
+{
+  "query": "authentication",
+  "type": "project",
+  "scope": "",
+  "scope_id": "",
+  "limit": 20
+}
+```
+
+| Field      | Type   | Required | Default | Description                    |
+|------------|--------|----------|---------|--------------------------------|
+| `query`    | string | yes      | --      | Search query                   |
+| `type`     | string | no       | `""`    | Filter by type                 |
+| `scope`    | string | no       | `""`    | Filter by scope                |
+| `scope_id` | string | no       | `""`    | Filter by scope ID             |
+| `limit`    | int    | no       | `20`    | Max results (capped at 50)     |
+
+**Response:** `200`
+
+```json
+{
+  "memories": [
+    {
+      "memory_id": "a1b2c3d4-e5f6-...",
+      "name": "auth_patterns",
+      "description": "Authentication architecture",
+      "type": "project",
+      "scope": "global",
+      "scope_id": "",
+      "content": "JWT tokens with HS256...",
+      "created": "2026-03-10T10:00:00",
+      "updated": "2026-03-12T14:30:00"
+    }
+  ],
+  "total": 1
+}
+```
+
+---
+
+### `DELETE /v1/api/memories/{name}`
+
+Delete a memory by name and scope.
+
+**Path parameters:**
+
+| Parameter | Type   | Description          |
+|-----------|--------|----------------------|
+| `name`    | string | Memory name          |
+
+**Query parameters:**
+
+| Parameter  | Type   | Required | Default    | Description         |
+|------------|--------|----------|------------|---------------------|
+| `scope`    | string | no       | `"global"` | Scope of the memory |
+| `scope_id` | string | no       | `""`       | Scope qualifier     |
+
+**Response (success):** `200`
+
+```json
+{"status": "ok", "name": "deployment_process"}
+```
+
+**Response (not found):** `404`
+
+```json
+{"error": "Memory 'deployment_process' not found"}
+```
+
+---
+
+## Console Admin API
+
+Four admin endpoints for cross-workstream memory management. All require the
+`admin.memories` permission.
+
+### `GET /v1/api/admin/memories`
+
+List memories across all scopes (no automatic scope resolution).
+
+**Query parameters:**
+
+| Parameter  | Type   | Required | Default | Description                  |
+|------------|--------|----------|---------|------------------------------|
+| `type`     | string | no       | `""`    | Filter by type               |
+| `scope`    | string | no       | `""`    | Filter by scope              |
+| `scope_id` | string | no       | `""`    | Filter by scope ID           |
+| `limit`    | int    | no       | `100`   | Max results (capped at 200)  |
+
+**Response:** `200`
+
+```json
+{
+  "memories": [
+    {
+      "memory_id": "a1b2c3d4-e5f6-...",
+      "name": "project_architecture",
+      "description": "Core architecture patterns",
+      "type": "project",
+      "scope": "global",
+      "scope_id": "",
+      "content": "The project uses...",
+      "created": "2026-03-10T10:00:00",
+      "updated": "2026-03-12T14:30:00"
+    }
+  ],
+  "total": 1
+}
+```
+
+---
+
+### `GET /v1/api/admin/memories/search`
+
+Search memories by query (uses query parameters, not POST body).
+
+**Query parameters:**
+
+| Parameter  | Type   | Required | Default | Description                   |
+|------------|--------|----------|---------|-------------------------------|
+| `q`        | string | yes      | --      | Search query                  |
+| `type`     | string | no       | `""`    | Filter by type                |
+| `scope`    | string | no       | `""`    | Filter by scope               |
+| `scope_id` | string | no       | `""`    | Filter by scope ID            |
+| `limit`    | int    | no       | `20`    | Max results (capped at 50)    |
+
+**Response:** `200` -- same schema as `GET /v1/api/admin/memories`.
+
+---
+
+### `GET /v1/api/admin/memories/{memory_id}`
+
+Get a single memory by ID.
+
+**Path parameters:**
+
+| Parameter   | Type   | Description            |
+|-------------|--------|------------------------|
+| `memory_id` | string | Memory UUID            |
+
+**Response (success):** `200`
+
+```json
+{
+  "memory_id": "a1b2c3d4-e5f6-...",
+  "name": "project_architecture",
+  "description": "Core architecture patterns",
+  "type": "project",
+  "scope": "global",
+  "scope_id": "",
+  "content": "The project uses...",
+  "created": "2026-03-10T10:00:00",
+  "updated": "2026-03-12T14:30:00"
+}
+```
+
+**Response (not found):** `404`
+
+```json
+{"error": "Memory not found"}
+```
+
+---
+
+### `DELETE /v1/api/admin/memories/{memory_id}`
+
+Delete a memory by ID. Records an audit event (`memory.delete`).
+
+**Path parameters:**
+
+| Parameter   | Type   | Description            |
+|-------------|--------|------------------------|
+| `memory_id` | string | Memory UUID            |
+
+**Response (success):** `200`
+
+```json
+{"status": "ok"}
+```
+
+**Response (not found):** `404`
+
+```json
+{"error": "Memory not found"}
+```
+
+---
+
+## SDK
+
+### Python
+
+The server SDK uses `mem_type` (not `type`) to avoid shadowing the Python
+builtin.
+
+```python
+from turnstone.sdk import TurnstoneServer
+
+with TurnstoneServer("http://localhost:8080", token="tok_xxx") as client:
+    # Save a memory
+    mem = client.save_memory(
+        "api_conventions",
+        "All endpoints use /v1/ prefix. JSON responses.",
+        description="API design patterns",
+        mem_type="project",
+        scope="global",
+    )
+    print(mem.memory_id)
+
+    # Search memories
+    results = client.search_memories("authentication", mem_type="project", limit=10)
+    for m in results.memories:
+        print(f"{m['name']}: {m['description']}")
+
+    # List memories
+    all_mems = client.list_memories(mem_type="feedback", limit=50)
+
+    # Delete a memory
+    client.delete_memory("api_conventions", scope="global")
+```
+
+Console admin SDK:
+
+```python
+from turnstone.sdk import TurnstoneConsole
+
+with TurnstoneConsole("http://localhost:9090", token="tok_xxx") as admin:
+    # List all memories (admin view, no scope auto-resolution)
+    result = admin.list_memories(scope="global", limit=100)
+
+    # Search
+    result = admin.search_memories("architecture", mem_type="project")
+
+    # Get by ID
+    mem = admin.get_memory("a1b2c3d4-e5f6-...")
+
+    # Delete by ID
+    admin.delete_memory("a1b2c3d4-e5f6-...")
+```
+
+### TypeScript
+
+```typescript
+import { TurnstoneServer } from "@turnstone/sdk";
+
+const client = new TurnstoneServer({
+  baseUrl: "http://localhost:8080",
+  token: "tok_xxx",
+});
+
+// Save a memory
+const mem = await client.saveMemory({
+  name: "api_conventions",
+  content: "All endpoints use /v1/ prefix. JSON responses.",
+  description: "API design patterns",
+  type: "project",
+  scope: "global",
+});
+
+// Search memories
+const results = await client.searchMemories({
+  query: "authentication",
+  type: "project",
+  limit: 10,
+});
+
+// List memories
+const all = await client.listMemories({ type: "feedback", limit: 50 });
+
+// Delete a memory
+await client.deleteMemory("api_conventions", { scope: "global" });
+```
+
+Console admin SDK:
+
+```typescript
+import { TurnstoneConsole } from "@turnstone/sdk";
+
+const admin = new TurnstoneConsole({
+  baseUrl: "http://localhost:9090",
+  token: "tok_xxx",
+});
+
+// List, search, get, delete by ID
+const mems = await admin.listMemories({ scope: "global" });
+const found = await admin.searchMemories({ q: "auth", limit: 20 });
+const one = await admin.getMemory("a1b2c3d4-e5f6-...");
+await admin.deleteMemory("a1b2c3d4-e5f6-...");
+```
+
+---
+
+## Storage
+
+Memories are stored in the `structured_memories` table (migration 013).
+The unique constraint on `(name, scope, scope_id)` ensures upsert semantics.
+The name is normalized on save: lowercased, hyphens and spaces replaced with
+underscores.
+
+## Architecture
+
+See [Memory Architecture diagram](diagrams/png/23-memory-architecture.png) for
+the full data flow covering the session tool path, API path, admin path, and
+BM25 relevance injection.

docs/sdk.md

@@ -0,0 +1,325 @@
+# Turnstone Client SDK
+
+> See also: [API Reference](api-reference.md) | [Architecture](architecture.md) | [SDK Class Diagram](diagrams/png/13-sdk-architecture.png)
+
+Typed HTTP client libraries for programmatic access to the turnstone server and console APIs. Available in Python (sync + async) and TypeScript.
+
+---
+
+## Python SDK
+
+The Python SDK is included in the `turnstone` package — no extra install required. It wraps the REST and SSE endpoints with typed methods that return Pydantic models directly.
+
+### Quick Start
+
+```python
+from turnstone.sdk import TurnstoneServer
+
+# Synchronous client — login with username/password
+with TurnstoneServer("http://localhost:8080") as client:
+    client.login(username="alice", password="s3cret")
+
+    # Create a workstream
+    ws = client.create_workstream(name="Analysis")
+
+    # Send a message and wait for the full response
+    result = client.send_and_wait("Summarize this codebase.", ws.ws_id)
+    print(result.content)
+
+    # Stream events in real time
+    for event in client.stream_events(ws.ws_id):
+        if event.type == "content":
+            print(event.text, end="", flush=True)
+
+    # Close when done
+    client.close_workstream(ws.ws_id)
+```
+
+Alternatively, authenticate with an API token:
+
+```python
+with TurnstoneServer("http://localhost:8080") as client:
+    client.login(token="ts_abc123...")
+    ws = client.create_workstream(name="CI run")
+    result = client.send_and_wait("Run the test suite.", ws.ws_id)
+```
+
+### Async Client
+
+```python
+import asyncio
+from turnstone.sdk import AsyncTurnstoneServer
+
+async def main():
+    async with AsyncTurnstoneServer("http://localhost:8080") as client:
+        await client.login(username="alice", password="s3cret")
+        ws = await client.create_workstream(name="demo")
+        async for event in client.stream_events(ws.ws_id):
+            if event.type == "content":
+                print(event.text, end="", flush=True)
+
+asyncio.run(main())
+```
+
+### Server Client API
+
+Both `TurnstoneServer` (sync) and `AsyncTurnstoneServer` (async) expose:
+
+| Category | Method | Returns |
+|----------|--------|---------|
+| **Workstreams** | `list_workstreams()` | `ListWorkstreamsResponse` |
+| | `dashboard()` | `DashboardResponse` |
+| | `create_workstream(*, name, model, auto_approve, ws_template)` | `CreateWorkstreamResponse` |
+| | `close_workstream(ws_id)` | `StatusResponse` |
+| **Chat** | `send(message, ws_id)` | `SendResponse` |
+| | `approve(*, ws_id, approved, feedback, always)` | `StatusResponse` |
+| | `plan_feedback(*, ws_id, feedback)` | `StatusResponse` |
+| | `command(*, ws_id, command)` | `StatusResponse` |
+| | `cancel(ws_id)` | `StatusResponse` |
+| **Streaming** | `stream_events(ws_id)` | `Iterator[ServerEvent]` |
+| | `stream_global_events()` | `Iterator[ServerEvent]` |
+| **High-level** | `send_and_wait(message, ws_id, *, timeout, on_event)` | `TurnResult` |
+| **Saved** | `list_saved_workstreams()` | `ListSavedWorkstreamsResponse` |
+| **Auth** | `login(username=..., password=...)` | `AuthLoginResponse` |
+| | `login(token="ts_xxx")` | `AuthLoginResponse` |
+| | `logout()` | `StatusResponse` |
+| | `auth_status()` | `AuthStatusResponse` |
+| **Health** | `health()` | `HealthResponse` |
+
+### Console Client API
+
+Both `TurnstoneConsole` (sync) and `AsyncTurnstoneConsole` (async) expose:
+
+| Category | Method | Returns |
+|----------|--------|---------|
+| **Cluster** | `overview()` | `ClusterOverviewResponse` |
+| | `nodes(*, sort, limit, offset)` | `ClusterNodesResponse` |
+| | `workstreams(*, state, node, search, sort, page, per_page)` | `ClusterWorkstreamsResponse` |
+| | `node_detail(node_id)` | `NodeDetailResponse` |
+| | `snapshot()` | `ClusterSnapshotResponse` |
+| | `create_workstream(*, node_id, name, model, initial_message, ws_template)` | `ConsoleCreateWsResponse` |
+| **Schedules** | `list_schedules()` | `ListSchedulesResponse` |
+| | `create_schedule(*, name, schedule_type, initial_message, ...)` | `ScheduleInfo` |
+| | `get_schedule(task_id)` | `ScheduleInfo` |
+| | `update_schedule(task_id, *, name=..., enabled=..., ...)` | `ScheduleInfo` |
+| | `delete_schedule(task_id)` | `StatusResponse` |
+| | `list_schedule_runs(task_id, *, limit=50)` | `ListScheduleRunsResponse` |
+| **WS Templates** | `list_ws_templates()` | `ListWsTemplatesResponse` |
+| | `create_ws_template(*, name, description, ...)` | `WsTemplateInfo` |
+| | `get_ws_template(template_id)` | `WsTemplateInfo` |
+| | `update_ws_template(template_id, *, name=..., enabled=..., ...)` | `WsTemplateInfo` |
+| | `delete_ws_template(template_id)` | `StatusResponse` |
+| | `list_ws_template_versions(template_id)` | `ListWsTemplateVersionsResponse` |
+| **Streaming** | `stream_cluster_events()` | `Iterator[ClusterEvent]` |
+| **Auth** | `login(username=..., password=...)` / `login(token="ts_xxx")` | `AuthLoginResponse` |
+| | `logout()` | `StatusResponse` |
+| **Health** | `health()` | `ConsoleHealthResponse` |
+
+### Event Types
+
+SSE events are deserialized into typed dataclasses. Use `event.type` to discriminate.
+
+**Per-workstream events** (from `stream_events(ws_id)`):
+
+| Type | Class | Key Fields |
+|------|-------|------------|
+| `connected` | `ConnectedEvent` | `model`, `model_alias`, `skip_permissions` |
+| `history` | `HistoryEvent` | `messages` |
+| `content` | `ContentEvent` | `text` |
+| `reasoning` | `ReasoningEvent` | `text` |
+| `tool_info` | `ToolInfoEvent` | `items` |
+| `approve_request` | `ApproveRequestEvent` | `items` |
+| `tool_result` | `ToolResultEvent` | `call_id`, `name`, `output` |
+| `tool_output_chunk` | `ToolOutputChunkEvent` | `call_id`, `chunk` |
+| `status` | `StatusEvent` | `prompt_tokens`, `total_tokens`, `pct`, `effort` |
+| `plan_review` | `PlanReviewEvent` | `content` |
+| `error` | `ErrorEvent` | `message` |
+| `info` | `InfoEvent` | `message` |
+| `stream_end` | `StreamEndEvent` | — |
+| `cancelled` | `CancelledEvent` | — |
+
+**Global events** (from `stream_global_events()`):
+
+| Type | Class | Key Fields |
+|------|-------|------------|
+| `ws_state` | `WsStateEvent` | `ws_id`, `state`, `tokens`, `activity` |
+| `ws_activity` | `WsActivityEvent` | `ws_id`, `activity`, `activity_state` |
+| `ws_rename` | `WsRenameEvent` | `ws_id`, `name` |
+| `ws_closed` | `WsClosedEvent` | `ws_id` |
+
+**Cluster events** (from `stream_cluster_events()`):
+
+| Type | Class | Key Fields |
+|------|-------|------------|
+| `node_joined` | `NodeJoinedEvent` | `node_id` |
+| `node_lost` | `NodeLostEvent` | `node_id` |
+| `cluster_state` | `ClusterStateEvent` | `ws_id`, `node_id`, `state`, `tokens` |
+| `ws_created` | `ClusterWsCreatedEvent` | `ws_id`, `node_id`, `name` |
+| `ws_closed` | `ClusterWsClosedEvent` | `ws_id` |
+| `ws_rename` | `ClusterWsRenameEvent` | `ws_id`, `name` |
+| `snapshot` | `ClusterSnapshotEvent` | `nodes`, `overview`, `timestamp` |
+
+### TurnResult
+
+The `send_and_wait()` method returns a `TurnResult` that aggregates the full response:
+
+```python
+result = client.send_and_wait("Hello", ws_id, timeout=60)
+result.content      # Full text response
+result.reasoning    # Chain-of-thought (if shown)
+result.tool_results # List of (tool_name, output) tuples
+result.errors       # Any error messages
+result.ok           # True if no errors and not timed out
+result.timed_out    # True if timeout expired
+```
+
+### Error Handling
+
+Non-2xx responses raise `TurnstoneAPIError`:
+
+```python
+from turnstone.sdk import TurnstoneServer, TurnstoneAPIError
+
+try:
+    client.send("hi", "bad_ws_id")
+except TurnstoneAPIError as e:
+    print(e.status_code)  # 404
+    print(e.message)      # "Unknown workstream"
+```
+
+---
+
+## TypeScript SDK
+
+Located at `sdk/typescript/`. Zero runtime dependencies for browsers; uses native `fetch` and `ReadableStream` for SSE parsing.
+
+### Quick Start
+
+```typescript
+import { TurnstoneServer } from "@turnstone/sdk";
+
+const client = new TurnstoneServer({ baseUrl: "http://localhost:8080" });
+
+// Login with username/password or API token
+await client.login({ username: "alice", password: "s3cret" });
+// or: await client.login({ token: "ts_abc123..." });
+
+// Create workstream and send message
+const ws = await client.createWorkstream({ name: "demo" });
+const result = await client.sendAndWait("Hello!", ws.ws_id);
+console.log(result.content);
+
+// Stream events
+for await (const event of client.streamEvents(ws.ws_id)) {
+  if (event.type === "content") {
+    process.stdout.write(event.text);
+  }
+}
+```
+
+### Console Client
+
+```typescript
+import { TurnstoneConsole } from "@turnstone/sdk";
+
+const client = new TurnstoneConsole({ baseUrl: "http://localhost:8090" });
+await client.login({ username: "alice", password: "s3cret" });
+
+const overview = await client.overview();
+console.log(`Nodes: ${overview.nodes}, Workstreams: ${overview.workstreams}`);
+
+// Stream cluster events
+for await (const event of client.clusterEvents()) {
+  console.log(event.type, event);
+}
+```
+
+### Type Safety
+
+All event types are modeled as a discriminated union:
+
+```typescript
+import { isContentEvent, isErrorEvent } from "@turnstone/sdk";
+import type { ServerEvent } from "@turnstone/sdk";
+
+function handleEvent(event: ServerEvent) {
+  if (isContentEvent(event)) {
+    // event is narrowed to ContentEvent
+    console.log(event.text);
+  } else if (isErrorEvent(event)) {
+    console.error(event.message);
+  }
+}
+```
+
+### Custom Fetch
+
+The client accepts a custom `fetch` implementation for testing or Node.js environments:
+
+```typescript
+const client = new TurnstoneServer({
+  baseUrl: "http://localhost:8080",
+  fetch: myCustomFetch,
+});
+```
+
+---
+
+## Architecture
+
+```
+turnstone/sdk/               Python SDK (sub-package)
+  _base.py                   Shared httpx async client, auth, error handling
+  _sync.py                   Background event loop for sync wrappers
+  _types.py                  TurnResult + TurnstoneAPIError
+  events.py                  27 SSE event dataclasses with type registry
+  server.py                  AsyncTurnstoneServer + TurnstoneServer
+  console.py                 AsyncTurnstoneConsole + TurnstoneConsole
+
+sdk/typescript/              TypeScript SDK (npm package)
+  src/base.ts                fetch wrapper, auth, SSE streaming
+  src/server.ts              TurnstoneServer class
+  src/console.ts             TurnstoneConsole class
+  src/events.ts              Discriminated union events + type guards
+  src/sse.ts                 ReadableStream SSE parser
+  src/types.ts               Request/response interfaces
+```
+
+The Python SDK reuses Pydantic models from `turnstone/api/` directly — no schema duplication. The TypeScript SDK has hand-written interfaces matching those models.
+
+Both SDKs follow the same design: typed methods for REST endpoints, async iterators for SSE streams, and a high-level `send_and_wait` method for simple request-response patterns.
+
+---
+
+## Authentication
+
+When auth is enabled on the server, the SDK handles JWT-based authentication automatically.
+
+### Login Flow
+
+There are two ways to authenticate:
+
+1. **Username + password** — calls `POST /v1/api/auth/login` with credentials. The server validates against the user database and returns a JWT.
+
+2. **API token** — calls `POST /v1/api/auth/login` with a `ts_`-prefixed token string. The server looks up the token, resolves the associated user, and returns a JWT.
+
+In both cases the server returns the JWT in the response body and as a `Set-Cookie` header. The SDK extracts the JWT and includes it as a `Bearer` token in the `Authorization` header on all subsequent requests.
+
+```python
+# Username + password
+client.login(username="alice", password="s3cret")
+
+# API token (created via admin API or turnstone-admin CLI)
+client.login(token="ts_abc123...")
+```
+
+### Token Lifecycle
+
+- JWTs have a configurable expiry (default: 24 hours).
+- `client.auth_status()` returns the current user identity and scopes without refreshing the token.
+- `client.logout()` clears the stored JWT from the client.
+- If a request returns 401, the SDK raises `TurnstoneAPIError` — the caller is responsible for re-authenticating.
+
+### Backward Compatibility
+
+The config-file token (`TURNSTONE_AUTH_TOKEN`) still works as a simple Bearer token for environments that do not use the user/JWT system. When the server receives a non-JWT Bearer token, it falls back to the legacy token check.

docs/security.md

@@ -0,0 +1,485 @@
+# Security and Authentication
+
+Turnstone uses a layered authentication system with three token types,
+hierarchical scopes, and a split architecture where the console manages
+credentials while individual server nodes validate JWTs locally.
+
+---
+
+## Token Types
+
+### Config-file tokens
+
+Static tokens defined in `config.toml` or the `TURNSTONE_AUTH_TOKEN`
+environment variable. Validated in-memory using `hmac.compare_digest`
+(timing-safe). Each token maps to a role that determines its scopes.
+
+```toml
+[[auth.tokens]]
+value = "tok_legacy"
+role = "full"     # full → {read, write, approve}
+```
+
+Role mappings: `"read"` → `{read}`, `"full"` → `{read, write, approve}`.
+
+Config tokens are sent directly as `Authorization: Bearer tok_legacy`
+on every request. No JWT exchange is needed.
+
+### API tokens
+
+Database-backed tokens prefixed with `ts_`. Created via the admin CLI
+(`turnstone-admin create-token`) or the console admin API. Stored as
+SHA-256 hashes — the raw token is shown exactly once at creation and
+never persisted in plaintext.
+
+```
+$ turnstone-admin create-token --user abc123 --scopes read,write --name "CI bot"
+Token created: ts_a1b2c3d4e5f6...
+(save this — it will not be shown again)
+```
+
+API tokens can be used directly as `Bearer ts_xxx` headers or exchanged
+for a JWT via the login endpoint.
+
+### JWTs
+
+Short-lived session tokens (24 hours by default). Issued after
+authenticating with username/password or by exchanging an API token.
+HS256-signed with a shared secret. Validated locally on every service
+node — no database call per request.
+
+Claims:
+
+| Claim | Description |
+|-------|-------------|
+| `sub` | User ID |
+| `scopes` | Comma-separated scope list (`read,write,approve`) |
+| `src` | Token source (`password`, `api_token`, `config`) |
+| `iss` | Issuer — always `turnstone` |
+| `aud` | Audience — `turnstone-server` or `turnstone-console` |
+| `iat` | Issued-at timestamp |
+| `exp` | Expiry timestamp |
+
+The `aud` claim prevents cross-service token reuse — a JWT issued for the
+console cannot be used to authenticate against a server node, and vice versa.
+Tokens without an `aud` claim are accepted during the rollout window when
+`audience` validation is not specified.
+
+---
+
+## Scope Model
+
+Scopes are hierarchical — higher scopes imply all lower ones.
+
+| Scope | Grants | Implies |
+|-------|--------|---------|
+| `read` | View workstreams, saved workstreams, history | — |
+| `write` | Send messages, create/close workstreams | `read` |
+| `approve` | Approve tool calls, admin endpoints | `read`, `write` |
+
+### Path-to-scope mapping
+
+| Method | Path pattern | Required scope |
+|--------|-------------|----------------|
+| GET | Any protected path | `read` |
+| POST | `/api/send`, `/api/plan`, `/api/command` | `write` |
+| POST | `/api/workstreams/new`, `/api/workstreams/close` | `write` |
+| POST | `/api/cluster/workstreams/new` | `write` |
+| POST | `/api/approve` | `approve` |
+| Any | `/api/admin/*` | `approve` |
+
+Public paths bypass authentication entirely: `/`, `/health`, `/metrics`,
+`/static/*`, `/shared/*`, `/docs`, `/openapi.json`, `/api/auth/login`,
+`/api/auth/logout`, `/api/auth/status`, `/api/auth/setup`.
+
+### RBAC (Granular Permissions)
+
+> See also: [Governance documentation](governance.md)
+
+Scopes provide coarse endpoint-level access control. For finer-grained
+enforcement, the governance layer adds 15 named permissions checked
+per-endpoint by `require_permission()`. Permissions are bundled into
+roles; users are assigned roles via the `user_roles` join table.
+
+At login, `_load_user_permissions()` aggregates all permissions from
+the user's assigned roles. `_permissions_to_scopes()` derives legacy
+scopes for backward compatibility (e.g., any `admin.*` permission
+implies the `approve` scope). The JWT carries both `scopes` and
+`permissions` claims.
+
+Three built-in roles are seeded by migration 008:
+
+| Role | Permissions |
+|------|-------------|
+| admin | All 15 permissions |
+| operator | read, write, workstreams.create, workstreams.close |
+| viewer | read |
+
+Custom roles can be created with any subset of the valid permissions.
+Role creation and update validate permissions against a static allowlist.
+Self-assignment is blocked, and assigning a role requires the caller to
+hold a superset of the target role's permissions.
+
+---
+
+## Login Flows
+
+### Username and password
+
+```
+POST /v1/api/auth/login
+Content-Type: application/json
+
+{"username": "admin", "password": "s3cret"}
+```
+
+Returns a JWT in the response body and sets an `HttpOnly` session cookie.
+
+### API token exchange
+
+```
+POST /v1/api/auth/login
+Content-Type: application/json
+
+{"token": "ts_a1b2c3d4e5f6..."}
+```
+
+The API token is hashed, looked up in the database, and exchanged for a
+JWT with the token's scopes. This is the recommended flow for SDKs and
+automated clients that need cookie-based sessions.
+
+### Config-file tokens (direct)
+
+Config tokens are validated per-request via `hmac.compare_digest`. No
+login exchange is needed — include the token as a `Bearer` header:
+
+```
+Authorization: Bearer tok_legacy
+```
+
+### First-time setup
+
+When no users exist in the database:
+
+1. `GET /v1/api/auth/status` returns `{"setup_required": true}`
+2. The UI presents a setup wizard
+3. `POST /v1/api/auth/setup` creates the first admin user and returns a
+   JWT in one atomic step (no auth required — this is a public endpoint)
+4. The endpoint returns `409 Conflict` if setup has already been completed
+   (i.e. users already exist in the database)
+5. Subsequent admin requests require `approve` scope
+
+The `/api/auth/setup` endpoint is available on both the server and
+console. It validates input before creating the user:
+
+- **username**: 1-64 ASCII characters
+- **display_name**: required (non-empty)
+- **password**: minimum 8 characters
+
+```
+POST /v1/api/auth/setup
+Content-Type: application/json
+
+{"username": "admin", "display_name": "Admin", "password": "strongpass"}
+```
+
+Response:
+
+```json
+{
+  "status": "ok",
+  "user_id": "u_abc123",
+  "username": "admin",
+  "role": "full",
+  "scopes": "approve,read,write",
+  "jwt": "eyJhbGciOiJIUzI1NiIs..."
+}
+```
+
+The response also sets an `HttpOnly` session cookie containing the JWT,
+so the browser is immediately authenticated after setup completes.
+
+---
+
+## Token Detection Order
+
+The auth middleware inspects the `Authorization: Bearer <token>` header
+and classifies the token:
+
+1. **Contains `.`** → JWT → validate HS256 signature and expiry
+2. **Starts with `ts_`** → API token → SHA-256 hash, database lookup
+3. **Otherwise** → config-file token → `hmac.compare_digest` against
+   each configured token
+
+If a session cookie is present and no `Authorization` header is sent,
+the cookie value is treated as a JWT (step 1).
+
+---
+
+## Password Storage
+
+Passwords are hashed with **bcrypt** using a random salt per password.
+Plaintext passwords are only accepted over HTTPS in production
+deployments.
+
+---
+
+## Cookie Security
+
+| Attribute | Value | Purpose |
+|-----------|-------|---------|
+| `HttpOnly` | `true` | Prevents JavaScript access |
+| `SameSite` | `Lax` | CSRF protection |
+| `Path` | `/` | Available to all routes |
+| `Max-Age` | 24 hours | Matches JWT expiry |
+| `Secure` | `true` (default) | Always set unless explicitly disabled for dev |
+
+---
+
+## JWT Configuration
+
+| Setting | Config key | Env var | Default |
+|---------|-----------|---------|---------|
+| Signing secret | `[auth] jwt_secret` | `TURNSTONE_JWT_SECRET` | Auto-generated ephemeral (warning logged) |
+| Expiry | `[auth] jwt_expiry_hours` | — | 24 hours |
+| Algorithm | — | — | HS256 (not configurable) |
+| Minimum secret length | — | — | 32 characters (warning if shorter) |
+
+All service nodes that need to validate JWTs must share the same signing
+secret. If no secret is configured, an ephemeral key is generated at
+startup and a warning is logged — JWTs will not survive restarts or work
+across nodes.
+
+The bridge and console **require** `TURNSTONE_JWT_SECRET` when no
+`--auth-token` is provided. They exit with an error if the secret is
+missing, since ephemeral secrets would silently break inter-service
+communication.
+
+---
+
+## Admin API Endpoints
+
+All admin endpoints require `approve` scope.
+
+### Users
+
+| Method | Path | Description |
+|--------|------|-------------|
+| POST | `/v1/api/admin/users` | Create user (username, display_name, password) |
+| GET | `/v1/api/admin/users` | List all users |
+| DELETE | `/v1/api/admin/users/{user_id}` | Delete user and cascade tokens |
+
+### API tokens
+
+| Method | Path | Description |
+|--------|------|-------------|
+| POST | `/v1/api/admin/users/{user_id}/tokens` | Create API token (returns raw value once) |
+| GET | `/v1/api/admin/users/{user_id}/tokens` | List tokens (prefix only, no hashes) |
+| DELETE | `/v1/api/admin/tokens/{token_id}` | Revoke token |
+
+---
+
+## CLI Administration
+
+The `turnstone-admin` command provides offline user and token management:
+
+```
+turnstone-admin create-user --username admin --name "Admin" [--password] [--token]
+turnstone-admin create-token --user <user_id> --scopes read,write --name "CI bot"
+turnstone-admin list-users
+turnstone-admin list-tokens
+turnstone-admin revoke-token <token_id>
+```
+
+When `--password` is omitted, the CLI prompts interactively. When
+`--token` is passed to `create-user`, an API token is created alongside
+the user and printed to stdout.
+
+---
+
+## Database Schema
+
+```sql
+CREATE TABLE users (
+    user_id    TEXT PRIMARY KEY,
+    username   TEXT NOT NULL UNIQUE,
+    display_name TEXT NOT NULL,
+    password_hash TEXT NOT NULL,
+    created    TEXT NOT NULL
+);
+
+CREATE TABLE api_tokens (
+    token_id     TEXT PRIMARY KEY,
+    token_hash   TEXT NOT NULL,       -- SHA-256 of raw token
+    token_prefix TEXT NOT NULL,       -- first 8 chars for display
+    user_id      TEXT NOT NULL REFERENCES users(user_id),
+    name         TEXT NOT NULL,
+    scopes       TEXT NOT NULL,       -- comma-separated
+    created      TEXT NOT NULL,
+    expires      TEXT                 -- nullable, ISO 8601
+);
+
+CREATE UNIQUE INDEX ix_api_tokens_hash ON api_tokens(token_hash);
+
+CREATE TABLE channel_users (
+    channel_type    TEXT NOT NULL,
+    channel_user_id TEXT NOT NULL,
+    user_id         TEXT NOT NULL REFERENCES users(user_id),
+    created         TEXT NOT NULL,
+    PRIMARY KEY (channel_type, channel_user_id)
+);
+```
+
+The `sessions` and `workstreams` tables have a nullable `user_id`
+column for attribution when auth is enabled.
+
+---
+
+## Revocation
+
+- **API tokens**: Deleting a token via the admin API or CLI prevents new
+  JWTs from being issued with that token. Existing JWTs derived from the
+  token remain valid until they expire (at most 24 hours).
+- **Config-file tokens**: Remove the token from `config.toml` and
+  restart the service. No JWTs are involved, so revocation is immediate.
+- **JWTs**: Cannot be individually revoked. Rely on short expiry (24h)
+  and revoke the underlying credential to prevent renewal.
+
+---
+
+## Architecture
+
+```
+Console (cluster-wide)              Server (per-node)
+┌──────────────────────┐           ┌──────────────────────┐
+│ User/Token CRUD (DB) │           │ JWT validation only  │
+│ Login: creds → JWT   │           │ (shared signing key) │
+│ Admin API endpoints  │           │ Config tokens: hmac  │
+│ Storage: users,      │           │ No auth DB needed    │
+│   api_tokens tables  │           │                      │
+└──────────────────────┘           └──────────────────────┘
+```
+
+The console owns the credential database and handles all user/token
+CRUD.  Individual server nodes only need the JWT signing secret to
+validate session tokens.  Config-file tokens are validated locally
+without any database.
+
+### Proxy auth forwarding
+
+When the console proxies requests to server nodes (via `/node/{id}/...`
+routes), it uses a dedicated **service proxy token** with
+`aud: turnstone-server` and `write` scope.  The user's console JWT
+(which has `aud: turnstone-console`) is **not** forwarded — it would be
+rejected by the server's audience validation.
+
+The proxy token is managed by a `ServiceTokenManager` that auto-rotates
+1-hour JWTs, refreshing at 80% of lifetime.  If `--auth-token` is
+provided, that static token is used instead.
+
+### Service-to-service authentication
+
+The bridge and console collector use `ServiceTokenManager` for
+auto-rotating JWTs when communicating with server nodes:
+
+| Service | Identity | Scope | Audience | Purpose |
+|---------|----------|-------|----------|---------|
+| Bridge | `bridge` | `approve` | `turnstone-server` | Tool approval proxy, message relay |
+| Console collector | `console-collector` | `read` | `turnstone-server` | Node health polling |
+| Console proxy | `console-proxy` | `write` | `turnstone-server` | Proxied API calls |
+| Channel notify | `system` | `write` | `turnstone-channel` | Notification delivery to channel gateway |
+
+Service tokens use 1-hour expiry with automatic refresh via
+`ServiceTokenManager`.  The bridge injects auth headers per-request via
+httpx event hooks to ensure rotated tokens are picked up on SSE
+reconnects.
+
+Note that the channel gateway uses a distinct JWT audience
+(`turnstone-channel`) from the server (`turnstone-server`) and console
+(`turnstone-console`).  A server-scoped JWT cannot authenticate to the
+channel gateway endpoint, and vice versa.
+
+---
+
+## Configuration Reference
+
+### config.toml
+
+```toml
+[auth]
+enabled = true
+jwt_secret = "your-secret-key-here"
+jwt_expiry_hours = 24
+
+[[auth.tokens]]
+value = "tok_legacy"
+role = "full"
+```
+
+### Environment variables
+
+| Variable | Description |
+|----------|-------------|
+| `TURNSTONE_AUTH_ENABLED=1` | Enable authentication |
+| `TURNSTONE_AUTH_TOKEN=tok_xxx` | Register a config-file token with `full` access |
+| `TURNSTONE_JWT_SECRET=xxx` | JWT signing secret (must match across nodes) |
+| `TURNSTONE_CORS_ORIGINS=` | CORS allowed origins (comma-separated; empty = same-origin only) |
+
+---
+
+## Login Rate Limiting
+
+The `/api/auth/login` endpoint is protected by a dedicated
+`LoginRateLimiter` (separate from the general API rate limiter).
+Limits are enforced per-IP and per-username with a sliding window:
+
+- **5 attempts** per **5-minute window** per key
+- Failed logins record against both `ip:{client_ip}` and `user:{username}`
+- Returns `429 Too Many Requests` with `Retry-After` header when exceeded
+- Successful logins do not consume the budget
+
+---
+
+## CORS Policy
+
+By default, no CORS headers are sent (same-origin only). To allow
+cross-origin requests, set `TURNSTONE_CORS_ORIGINS`:
+
+```bash
+# Allow specific origins
+TURNSTONE_CORS_ORIGINS=https://app.example.com,https://admin.example.com
+
+# Allow all origins (development only)
+TURNSTONE_CORS_ORIGINS=*
+```
+
+When the variable is empty or unset, the CORS middleware is not added
+and browsers enforce same-origin policy.
+
+---
+
+## Security Properties
+
+- **Timing-safe comparison** for config-file tokens via
+  `hmac.compare_digest` — no timing side-channel.
+- **Hash-based lookup** for API tokens — the database stores only
+  SHA-256 hashes, eliminating timing attacks on token comparison.
+- **Local JWT validation** — no network call or database query needed
+  per request on server nodes.
+- **One-time display** of raw API tokens at creation. The plaintext is
+  never stored; `token_hash` never appears in API responses or logs.
+- **Structured logging audit trail** — `ctx_user_id` is set on every
+  authenticated request and injected into all log events.
+- **Scope enforcement** at the middleware layer before any handler
+  executes. Path-to-scope mapping is defined statically.
+- **JWT audience isolation** — server and console JWTs have distinct
+  `aud` claims, preventing cross-service token reuse.
+- **Login brute-force protection** — per-IP and per-username rate
+  limiting on the login endpoint.
+- **Secure cookies by default** — `Secure` flag set unconditionally;
+  24-hour max-age matches JWT expiry.
+- **CORS restriction** — no CORS headers by default (same-origin only).
+- **Service JWT auto-rotation** — 1-hour expiry with transparent
+  refresh, eliminating long-lived static tokens for inter-service auth.
+- **Secret strength validation** — warning logged when JWT secret is
+  shorter than 32 characters.

docs/settings.md

@@ -0,0 +1,353 @@
+# System Settings
+
+> See also: [Settings Architecture diagram](diagrams/png/24-settings-architecture.png)
+
+The system settings feature provides database-backed configuration for server
+nodes. Settings are stored in the `system_settings` table and managed through
+the admin API or console Settings tab. This replaces `config.toml` for
+non-bootstrap settings on server entry points, while the CLI continues to read
+`config.toml` directly.
+
+## Overview
+
+Settings follow a typed registry pattern: every storable setting has a
+`SettingDef` entry in `settings_registry.py` with type, default, description,
+validation constraints, and a `restart_required` flag. Unknown keys are rejected
+at the API boundary.
+
+At runtime, `ConfigStore` loads all settings from storage into an in-memory
+cache. Reads are lock-free dict lookups on an immutable snapshot. Writes acquire
+a lock, persist to storage, and swap the cache atomically.
+
+---
+
+## Precedence
+
+Settings resolution differs between entry points:
+
+| Entry point | Chain |
+|-------------|-------|
+| **Server** (`turnstone-server`, `turnstone-bridge`) | CLI flag > ConfigStore > registry default |
+| **CLI** (`turnstone`) | CLI flag > config.toml > argparse default |
+
+The server's `apply_config()` ignores config.toml sections that overlap with
+ConfigStore. A startup warning is logged for each overlapping key, directing
+users to the admin Settings API.
+
+---
+
+## Bootstrap vs ConfigStore
+
+**Bootstrap settings** are required before storage is available (database
+connection, Redis, auth secrets, server bind address). These stay in
+`config.toml` and environment variables.
+
+| Category | Section | Where |
+|----------|---------|-------|
+| API credentials | `[api]` | config.toml / env |
+| Database | `[database]` | config.toml / env |
+| Redis | `[redis]` | config.toml / env |
+| Auth | `[auth]` | config.toml / env |
+| Bridge identity | `[bridge]` | config.toml / env |
+| Console bind | `[console]` | config.toml / env |
+
+**ConfigStore settings** (~40 settings) are loaded from the database after
+storage initialization:
+
+| Section | Settings |
+|---------|----------|
+| `model` | name, temperature, max_tokens, reasoning_effort, context_window |
+| `session` | instructions, retention_days, compact_max_tokens, auto_compact_pct |
+| `tools` | timeout, truncation, agent_max_turns, skip_permissions, search, search_threshold, search_max_results |
+| `server` | workstream_idle_timeout, max_workstreams |
+| `mcp` | config_path, refresh_interval |
+| `ratelimit` | enabled, requests_per_second, burst |
+| `health` | backend_probe_interval, backend_probe_timeout, circuit_breaker_threshold, circuit_breaker_cooldown |
+| `judge` | enabled, model, provider, base_url, api_key, confidence_threshold, max_context_ratio, timeout, read_only_tools |
+| `memory` | relevance_k, fetch_limit, max_content, nudge_cooldown, nudges |
+
+Settings are addressed by dotted key (e.g. `memory.relevance_k`). Each has a
+declared type (`int`, `float`, `str`, `bool`), optional `min_value`/`max_value`
+range, optional `choices` list, and an `is_secret` flag.
+
+---
+
+## Storage
+
+The `system_settings` table (migration 015) stores settings as JSON-encoded
+values with a composite primary key of `(key, node_id)`:
+
+| Column | Type | Description |
+|--------|------|-------------|
+| `key` | text | Dotted setting key (e.g. `model.temperature`) |
+| `value` | text | JSON-encoded value |
+| `node_id` | text | Node ID for per-node overrides (empty string = global) |
+| `is_secret` | int | 1 if the setting contains secrets |
+| `changed_by` | text | Username of last editor |
+| `created` | text | ISO timestamp |
+| `updated` | text | ISO timestamp |
+
+Per-node overrides layer on top of global settings. When `ConfigStore` loads,
+it fetches global settings first, then overlays per-node values.
+
+---
+
+## Admin API
+
+Four endpoints on the **console** server, all requiring the `admin.settings`
+permission.
+
+### `GET /v1/api/admin/settings`
+
+List all settings with their effective values, defaults, and metadata.
+
+**Response:** `200`
+
+```json
+{
+  "settings": [
+    {
+      "key": "model.temperature",
+      "value": 0.7,
+      "source": "storage",
+      "type": "float",
+      "description": "Sampling temperature",
+      "section": "model",
+      "is_secret": false,
+      "node_id": "",
+      "changed_by": "admin",
+      "updated": "2026-03-14T10:00:00",
+      "restart_required": false
+    }
+  ]
+}
+```
+
+---
+
+### `GET /v1/api/admin/settings/schema`
+
+Return the full registry catalog (all defined settings with metadata). Useful
+for building dynamic admin UIs.
+
+**Response:** `200`
+
+```json
+{
+  "schema": [
+    {
+      "key": "model.temperature",
+      "type": "float",
+      "default": 0.5,
+      "description": "Sampling temperature",
+      "section": "model",
+      "is_secret": false,
+      "min_value": 0.0,
+      "max_value": 2.0,
+      "choices": null,
+      "restart_required": false
+    }
+  ]
+}
+```
+
+---
+
+### `PUT /v1/api/admin/settings/{key}`
+
+Update a setting. The value is validated against the registry (type coercion,
+range, choices). Secret settings (`is_secret=true`) cannot be written via the
+API -- they must be configured via config.toml or environment variables.
+
+**Path parameters:**
+
+| Parameter | Type   | Description |
+|-----------|--------|-------------|
+| `key`     | string | Dotted setting key (e.g. `model.temperature`) |
+
+**Request body:**
+
+```json
+{
+  "value": 0.7,
+  "node_id": ""
+}
+```
+
+| Field     | Type   | Required | Default | Description |
+|-----------|--------|----------|---------|-------------|
+| `value`   | any    | yes      | --      | New value (type-coerced against registry) |
+| `node_id` | string | no       | `""`    | Node ID for per-node override |
+
+**Response (success):** `200`
+
+```json
+{
+  "key": "model.temperature",
+  "value": 0.7,
+  "source": "storage",
+  "type": "float",
+  "description": "Sampling temperature",
+  "section": "model",
+  "is_secret": false,
+  "node_id": "",
+  "changed_by": "admin",
+  "updated": "",
+  "restart_required": false
+}
+```
+
+**Errors:**
+
+| Status | Condition |
+|--------|-----------|
+| 400    | Unknown key, invalid value, type mismatch, out of range |
+| 403    | Secret setting (must use config.toml or env) |
+
+---
+
+### `DELETE /v1/api/admin/settings/{key}`
+
+Reset a setting to its registry default by removing it from storage.
+
+**Path parameters:**
+
+| Parameter | Type   | Description |
+|-----------|--------|-------------|
+| `key`     | string | Dotted setting key |
+
+**Query parameters:**
+
+| Parameter | Type   | Required | Default | Description |
+|-----------|--------|----------|---------|-------------|
+| `node_id` | string | no       | `""`    | Node ID (empty = global) |
+
+**Response (success):** `200`
+
+```json
+{"status": "ok", "key": "model.temperature", "default": 0.5}
+```
+
+**Response (not found):** `404`
+
+```json
+{"error": "Setting 'model.temperature' has no stored value"}
+```
+
+---
+
+## Secret Settings
+
+Settings with `is_secret=True` (currently only `judge.api_key`) are blocked
+from the write API with a `403` response. This prevents accidental exposure
+through the admin UI or audit logs. Secret settings must be configured via
+`config.toml` or environment variables.
+
+The list endpoint masks secret values: stored secrets appear as `"***"`
+rather than their actual value.
+
+---
+
+## Hot Reload
+
+`ConfigStore` caches all settings in memory for fast, lock-free reads. To
+refresh the cache after external changes (e.g. direct database edits or
+cluster-wide propagation):
+
+```
+POST /v1/api/_internal/config-reload
+```
+
+This triggers `ConfigStore.reload()`, which re-reads all settings from storage
+and atomically swaps the cache. The `version` counter increments on every
+reload.
+
+**Behavior after reload:**
+
+- New workstreams pick up updated values immediately (via `session_factory`)
+- Existing sessions keep their frozen configuration (settings are captured at
+  workstream creation time, not read on every turn)
+- Settings marked `restart_required=True` need a server restart to take effect
+
+---
+
+## Migration from config.toml
+
+On startup, `warn_migrated_settings()` scans `config.toml` for keys that are
+now managed by ConfigStore. Each overlap produces a warning:
+
+```
+WARNING config.toml [model] temperature is now managed via Settings API —
+this value will be ignored. Use the admin Settings tab or
+PUT /v1/api/admin/settings/model.temperature to configure.
+```
+
+To migrate:
+
+1. Note the values from `config.toml` for sections that overlap with ConfigStore
+2. Use `PUT /v1/api/admin/settings/{key}` or the console Settings tab to set
+   each value
+3. Remove the migrated sections from `config.toml`
+4. Restart the server to verify no warnings
+
+---
+
+## SDK
+
+### Python
+
+```python
+from turnstone.sdk import TurnstoneConsole
+
+with TurnstoneConsole("http://localhost:9090", token="tok_xxx") as admin:
+    # List all settings with effective values
+    result = admin.list_settings()
+    for s in result["settings"]:
+        print(f"{s['key']} = {s['value']} (source: {s['source']})")
+
+    # Get the schema catalog
+    schema = admin.get_settings_schema()
+
+    # Update a setting
+    admin.update_setting("model.temperature", value=0.7)
+
+    # Update with per-node override
+    admin.update_setting("model.temperature", value=0.3, node_id="node-2")
+
+    # Reset to default
+    admin.delete_setting("model.temperature")
+```
+
+### TypeScript
+
+```typescript
+import { TurnstoneConsole } from "@turnstone/sdk";
+
+const admin = new TurnstoneConsole({
+  baseUrl: "http://localhost:9090",
+  token: "tok_xxx",
+});
+
+// List all settings
+const result = await admin.listSettings();
+for (const s of result.settings) {
+  console.log(`${s.key} = ${s.value} (source: ${s.source})`);
+}
+
+// Get schema catalog
+const schema = await admin.getSettingsSchema();
+
+// Update a setting
+await admin.updateSetting("model.temperature", { value: 0.7 });
+
+// Reset to default
+await admin.deleteSetting("model.temperature");
+```
+
+---
+
+## Architecture
+
+See [Settings Architecture diagram](diagrams/png/24-settings-architecture.png)
+for the full data flow covering server startup, admin API writes, hot reload,
+and settings precedence.

docs/tools.md

@@ -1,6 +1,6 @@
 # Tools Reference
 
-turnstone exposes 14 built-in tools plus any number of external MCP tools to the
+turnstone exposes 17 built-in tools plus any number of external MCP tools to the
 LLM via the OpenAI function-calling interface. Built-in tools are defined as JSON
 files under `turnstone/tools/` and loaded at startup by `turnstone/core/tools.py`.
 MCP tools are discovered from configured MCP servers at startup by
@@ -46,11 +46,12 @@ schema plus turnstone-specific metadata keys:
 
 | Name                | Description |
 |---------------------|-------------|
-| `TOOLS`             | All 14 tool definitions (sent to the model). |
+| `TOOLS`             | All 17 tool definitions (sent to the model). |
 | `AGENT_TOOLS`       | Tools with `agent: true` -- available to plan sub-agents. Read-only tools. |
 | `TASK_AGENT_TOOLS`  | Tools with `task_agent: true` -- available to task sub-agents. Includes write operations. |
 | `AGENT_AUTO_TOOLS`  | Set of tool names with `auto_approve: true` -- no user confirmation needed. |
 | `TASK_AUTO_TOOLS`   | Same as `AGENT_AUTO_TOOLS` (identical filter). |
+| `BUILTIN_TOOL_NAMES`| Frozenset of all 17 built-in tool names. Used by tool search to distinguish always-on tools from deferrable MCP tools. |
 | `PRIMARY_KEY_MAP`   | Dict mapping tool name to its `primary_key` parameter name. |
 
 ---
@@ -68,6 +69,9 @@ Tool execution follows a three-phase pipeline inside `ChatSession._execute_tools
 - Parses the JSON arguments (with fallback for malformed JSON).
 - If JSON parsing fails entirely, uses `PRIMARY_KEY_MAP` to map a bare string
   to the correct parameter.
+- Dispatches to the matching `_prepare_{func_name}()` handler. There are 17
+  built-in tools plus `tool_search` (synthetic, client-side BM25 fallback) and
+  the generic `_prepare_mcp_tool()` handler for MCP tools.
 - Validates arguments and builds a preview dict containing:
   - `call_id`, `func_name`, `header`, `preview` (for display)
   - `needs_approval` (bool)
@@ -110,9 +114,9 @@ Each item's `execute` callable is invoked:
 - `read_file` -- reads files, no side effects
 - `search` -- grep-style search, no side effects
 - `man` -- reads man pages, no side effects
-- `remember` -- writes to persistent memory database (lightweight, always auto-approved)
-- `recall` -- reads from persistent memory database
-- `forget` -- deletes from persistent memory database (lightweight, always auto-approved)
+- `memory` -- structured persistent memory (save/search/delete/list)
+- `recall` -- searches conversation history
+- `notify` -- sends notifications to linked channels (time-sensitive, auto-approved for urgency)
 
 **Requires user confirmation** (write operations, network access, side effects):
 - `bash` -- arbitrary command execution
@@ -159,9 +163,11 @@ Every tool defines a `primary_key`. The mapping is:
 | `web_search` | `query`     |
 | `task`       | `prompt`    |
 | `plan`       | `prompt`    |
-| `remember`   | `key`       |
+| `memory`     | `name`      |
 | `recall`     | `query`     |
-| `forget`     | `key`       |
+| `notify`     | `message`   |
+| `read_resource` | `uri`    |
+| `use_prompt` | `name`     |
 
 ---
 
@@ -183,15 +189,17 @@ Execute a bash command and return stdout + stderr.
 
 ### read_file
 
-Read the contents of a file, returning numbered lines.
+Read the contents of a file, returning numbered lines for text files or
+base64-encoded image data for supported image formats.
 
 | Parameter | Type    | Required | Description |
 |-----------|---------|----------|-------------|
 | `path`    | string  | yes      | Absolute or relative file path. |
-| `offset`  | integer | no       | Line number to start from (1-based, default: 1). |
-| `limit`   | integer | no       | Maximum number of lines to read. Omit for full file. |
+| `offset`  | integer | no       | Line number to start from (1-based, default: 1). Text files only. |
+| `limit`   | integer | no       | Maximum number of lines to read. Omit for full file. Text files only. |
 
-- **What it does**: Reads the file and returns content with line numbers. Must be called before `edit_file` on the same path (the session tracks which files have been read).
+- **What it does**: For text files, reads and returns content with line numbers. For image files (PNG, JPEG, GIF, WebP, BMP, TIFF, ICO), returns image data as multi-part content when the model supports vision, or a text description when it does not. SVG files are read as text. Images larger than 4 MB are rejected. Must be called before `edit_file` on the same path (the session tracks which files have been read).
+- **Vision support**: Controlled by `ModelCapabilities.supports_vision`. All commercial OpenAI and Anthropic models have vision enabled. Local models (vLLM, llama.cpp, NIM) default to off — enable via `[models.*.capabilities] supports_vision = true` in config.toml.
 - **Auto-approve**: Yes.
 - **Agent availability**: `agent` and `task_agent`.
 
@@ -302,8 +310,11 @@ Search the web using a text query.
 | `max_results` | integer | no       | Max results to return (default 5, max 20). |
 | `topic`       | string  | no       | Search topic: `general`, `news`, or `finance` (default `general`). |
 
-- **What it does**: Searches the web via the Tavily API and returns ranked results with titles, URLs, and content snippets.
-- **Auto-approve**: No -- requires user confirmation (makes network requests).
+- **What it does**: Searches the web and returns ranked results with titles, URLs, and content snippets. Uses provider-native search when available:
+  - **Anthropic**: Replaced at the API boundary with Anthropic's `web_search_20250305` server-side tool. Claude decides when to search; the API executes it and returns results with citations inline. No Tavily key needed.
+  - **OpenAI search models** (`gpt-5-search-api`): Replaced with `web_search_options` parameter. The model always searches and returns `url_citation` annotations.
+  - **Local/vLLM models**: Falls back to the Tavily API. Requires `tavily_key` in `config.toml` or `$TAVILY_API_KEY`.
+- **Auto-approve**: Yes (auto-approved for all tool dispatch paths).
 - **Agent availability**: `agent` and `task_agent`.
 
 ---
@@ -332,7 +343,7 @@ Plan before implementing -- an autonomous agent explores the codebase and writes
 |-----------|--------|----------|-------------|
 | `prompt`  | string | yes      | What to plan -- the goal, constraints, and scope. |
 
-- **What it does**: Spawns a planning sub-agent with `AGENT_TOOLS` (read-only tools: `read_file`, `search`, `math`, `man`, `web_fetch`, `web_search`). The agent explores the codebase and writes a structured plan to `.plan-<session_id>.md` (unique per session, so concurrent workstreams never collide). If the `plan` tool has been called before in the same session, the prior plan is passed to the agent as context so it refines rather than restarts. After completion, the user is prompted to review and can accept, reject, or annotate the plan.
+- **What it does**: Spawns a planning sub-agent with `AGENT_TOOLS` (read-only tools: `read_file`, `search`, `math`, `man`, `web_fetch`, `web_search`). The agent explores the codebase and writes a structured plan to `.plan-<ws_id>.md` (unique per workstream, so concurrent workstreams never collide). If the `plan` tool has been called before in the same session, the prior plan is passed to the agent as context so it refines rather than restarts. After completion, the user is prompted to review and can accept, reject, or annotate the plan.
 - **Auto-approve**: No -- requires user confirmation, plus post-execution review gate.
 - **Agent availability**: Not available to sub-agents (top-level only).
 
@@ -340,16 +351,22 @@ Plan before implementing -- an autonomous agent explores the codebase and writes
 
 ## Memory
 
-### remember
+### memory
 
-Save a persistent memory that persists across sessions.
+Structured persistent memory across sessions with typed, scoped entries.
 
-| Parameter | Type   | Required | Description |
-|-----------|--------|----------|-------------|
-| `key`     | string | yes      | Short identifier (e.g. `user_name`). |
-| `value`   | string | yes      | Content to remember. |
+| Parameter     | Type    | Required | Description |
+|---------------|---------|----------|-------------|
+| `action`      | string  | yes      | `save`, `search`, `delete`, or `list`. |
+| `name`        | string  | save/delete | Short snake_case identifier for the memory. |
+| `content`     | string  | save     | Memory content to store. |
+| `description` | string  | no       | Short description for relevance matching (recommended for `save`). |
+| `type`        | string  | no       | Memory type: `user`, `project`, `feedback`, or `reference`. Default: `project`. |
+| `scope`       | string  | no       | Memory scope: `global`, `workstream`, or `user`. Default: `global`. |
+| `query`       | string  | search   | Search query for finding memories. |
+| `limit`       | integer | no       | Max results for `search` or `list`. Default: 20. |
 
-- **What it does**: Stores a key-value pair in the SQLite memory database. Memories persist across sessions and are included in the system prompt on startup.
+- **What it does**: Manages structured persistent memories in the database. Memories persist across sessions, have a type classification (user preferences, project knowledge, feedback, reference material) and a scope (global across all workstreams, private to a workstream, or following a user). Relevant memories are included in the system prompt on startup.
 - **Auto-approve**: Yes.
 - **Agent availability**: Not available to sub-agents (top-level only).
 
@@ -357,30 +374,118 @@ Save a persistent memory that persists across sessions.
 
 ### recall
 
-Search memories and past conversations.
+Search conversation history for past messages and tool results.
 
 | Parameter | Type    | Required | Description |
 |-----------|---------|----------|-------------|
-| `query`   | string  | no       | Search term or phrase. Omit to list all memories. |
-| `limit`   | integer | no       | Max conversation results to return (default 20). |
+| `query`   | string  | yes      | Search term or phrase to find in conversation history. |
+| `limit`   | integer | no       | Max results to return (default 20). |
 
-- **What it does**: With no query, lists all saved memories. With a query, searches both the memory store and conversation history using FTS5 full-text search.
+- **What it does**: Searches conversation history across sessions using FTS5 full-text search. Returns matching messages, tool calls, and tool results with timestamps and workstream context.
 - **Auto-approve**: Yes.
 - **Agent availability**: Not available to sub-agents (top-level only).
 
 ---
 
-### forget
+## Notifications
 
-Remove a persistent memory by key.
+### notify
 
-| Parameter | Type   | Required | Description |
-|-----------|--------|----------|-------------|
-| `key`     | string | yes      | The memory key to remove (e.g. `user_name`). |
+Send a notification to a user or channel on an external platform.
 
-- **What it does**: Deletes the memory entry with the given key from the SQLite database.
-- **Auto-approve**: Yes.
-- **Agent availability**: Not available to sub-agents (top-level only).
+| Parameter      | Type   | Required | Description |
+|----------------|--------|----------|-------------|
+| `message`      | string | yes      | Notification content (plain text, max 2000 chars). |
+| `username`     | string | no       | Turnstone username — sends to all linked channels. |
+| `channel_type` | string | no       | Platform for direct targeting (`discord`). |
+| `channel_id`   | string | no       | Platform-specific channel or user ID for direct targeting. |
+| `title`        | string | no       | Optional short title (rendered as bold prefix). |
+
+Provide either `username` for user-based targeting or `channel_type` +
+`channel_id` for direct targeting. Do not combine both.
+
+- **What it does**: Sends a notification via the channel gateway's HTTP endpoint (`POST /v1/api/notify`). The server queries the `services` table for healthy channel gateways, authenticates with a service JWT (`aud: turnstone-channel`), and delivers to the first healthy gateway. On failure, retries up to 2 additional times with backoff (1s, 3s). Rate-limited to 5 notifications per turn (counter only increments on success).
+- **Auto-approve**: Yes — notifications are time-sensitive and auto-approved so the model can alert users urgently.
+- **Agent availability**: `agent` and `task_agent`.
+
+> See [Channel Integrations: Notifications](channels.md#notifications)
+> for the full delivery flow, service registry details, and security
+> measures.
+
+---
+
+### watch
+
+Set up periodic polling of a shell command within the current workstream.
+Results are injected back into the conversation as synthetic user messages,
+triggering the model to respond and act. Use for monitoring CI/CD pipelines,
+PR reviews, deployments, file changes, etc.
+
+| Parameter   | Type    | Required | Description |
+|-------------|---------|----------|-------------|
+| `action`    | string  | yes      | `create`, `list`, or `cancel`. |
+| `command`   | string  | create   | Shell command to poll periodically. |
+| `poll_every`| string  | no       | Poll interval as duration (`30s`, `5m`, `1h`). Default: `5m`. |
+| `stop_on`   | string  | no       | Python expression for stop condition (see below). Omit for change detection. |
+| `name`      | string  | create   | Human-readable watch name (e.g. `pr-review`). Used as identifier for cancel. |
+| `max_polls` | integer | no       | Max poll cycles before auto-cancel. Default: 100. |
+
+**Actions:**
+
+- `create` — Start a new watch. Requires approval (same as bash — runs shell
+  commands). Persists to the `watches` table; the server-level `WatchRunner`
+  daemon polls every 15 seconds for due watches.
+- `list` — Show all active watches in this workstream. Auto-approved.
+- `cancel` — Stop a watch by name or ID prefix. Auto-approved.
+
+**Stop condition DSL** — The `stop_on` parameter accepts a Python expression
+evaluated after each poll. Available variables:
+
+| Variable      | Type       | Description |
+|---------------|------------|-------------|
+| `output`      | `str`      | stdout (+stderr) of the command. |
+| `data`        | `Any`      | `json.loads(output)`, or `None` if not valid JSON. |
+| `exit_code`   | `int`      | Process exit code. |
+| `prev_output` | `str|None` | Previous poll's stdout (`None` on first poll). |
+| `changed`     | `bool`     | `True` if output differs from previous poll. |
+
+Safe builtins: `len`, `str`, `int`, `float`, `bool`, `abs`, `min`, `max`,
+`any`, `all`, `isinstance`, `sorted`. No `import`, `open`, `exec`, or
+`eval`. Security model: equivalent to `bash` — the model already has shell
+access.
+
+**Examples:**
+```
+data["state"] == "MERGED"
+"error" in output
+exit_code != 0
+changed and "ready" in output.lower()
+data.get("mergedAt") is not None
+```
+
+**Lifecycle:**
+
+1. Model calls `watch(action="create", ...)` — persisted to SQLite.
+2. `WatchRunner` daemon polls for due watches every 15s.
+3. Each poll runs the command, evaluates the condition.
+4. When the condition fires (or max polls reached), the result is injected
+   as a synthetic user message and the watch auto-cancels.
+5. If the workstream was evicted, it is restored before injection.
+6. Watches survive server restart (overdue watches fire once on recovery).
+
+**Constraints:**
+
+- Max 5 active watches per workstream.
+- Poll interval: 10s–24h.
+- Output truncated at 64 KB.
+- Max 5 consecutive watch dispatches per worker thread (depth guard).
+- Duplicate names rejected within the same workstream.
+
+- **Auto-approve**: `create` requires approval; `list` and `cancel` are auto-approved.
+- **Agent availability**: Main session only — not available to plan/task sub-agents.
+
+> See [Watch Architecture](diagrams/png/18-watch-architecture.png) for the
+> full poll → evaluate → dispatch flow.
 
 ---
 
@@ -399,14 +504,90 @@ Remove a persistent memory by key.
 | `web_search` | Info       | No           | Yes   | Yes        | `query`     |
 | `task`       | Agent      | No           | No    | No         | `prompt`    |
 | `plan`       | Agent      | No           | No    | No         | `prompt`    |
-| `remember`   | Memory     | Yes          | No    | No         | `key`       |
+| `memory`     | Memory     | Yes          | No    | No         | `name`      |
 | `recall`     | Memory     | Yes          | No    | No         | `query`     |
-| `forget`     | Memory     | Yes          | No    | No         | `key`       |
+| `notify`     | Notify     | Yes          | Yes   | Yes        | `message`   |
+| `watch`      | Monitor    | No (create)  | No    | No         | `command`   |
+| `read_resource`| MCP      | No           | Yes   | Yes        | `uri`       |
+| `use_prompt` | MCP        | No           | Yes   | Yes        | `name`      |
+| `tool_search`| Search     | Yes          | No    | No         | `query`     |
+
+---
+
+## Dynamic Tool Search
+
+When many MCP tools are connected, the total tool count can grow large enough to
+consume significant context window tokens and reduce model accuracy. Dynamic tool
+search addresses this by deferring tools the model is unlikely to need on the
+current turn and letting it search for them on demand.
+
+### Three-tier approach
+
+Tool search uses the best available mechanism for each provider:
+
+1. **Anthropic (native)** -- Models that support it receive `defer_loading: true`
+   on deferred tool definitions plus the `tool_search_tool_bm25_20251119` server-side
+   search tool. Anthropic's API handles search and expansion transparently.
+
+2. **OpenAI GPT-5.4+ (native)** -- Models with hosted tool search receive
+   `defer_loading: true` on deferred definitions. The API handles search internally.
+
+3. **vLLM / llama.cpp / NIM (client-side BM25)** -- A synthetic `tool_search`
+   function tool is injected into the tool list. When the model calls it,
+   `_exec_tool_search()` runs a pure-Python BM25 index over tool names and
+   descriptions, then expands the matched tools into the visible set.
+
+### Configuration
+
+Tool search is configured in `config.toml` under the `[tools]` section:
+
+```toml
+[tools]
+search = "auto"           # "auto", "on", or "off"
+search_threshold = 20     # minimum total tool count to activate
+search_max_results = 5    # max tools returned per search call
+```
+
+CLI flags override the config file:
+
+- `--tool-search {auto,on,off}` -- force tool search on or off, or let turnstone
+  decide based on threshold (default: `auto`).
+- `--tool-search-threshold N` -- minimum tool count to activate (default: 20).
+- `--tool-search-max-results N` -- max results per search (default: 5).
+
+### How it works
+
+1. **Threshold check**: At session startup, `ToolSearchManager.should_activate()`
+   counts total tools (built-in + MCP). If the count is below the threshold, tool
+   search stays off and all tools are sent to the model directly.
+
+2. **Partitioning**: When active, tools are split into two sets:
+   - **Always-on** -- the 17 built-in tools (members of `BUILTIN_TOOL_NAMES`).
+     These are always visible to the model.
+   - **Deferred** -- all MCP tools. These are not sent in the tool list unless
+     the model searches for them.
+
+3. **Search and expand**: When the model calls `tool_search` (client-side) or the
+   provider's native search returns results, the matched tools are added to the
+   visible set via `expand_visible()`. Once expanded, a tool stays visible for
+   the remainder of the session.
+
+4. **Multi-turn persistence**: Expanded tools are never removed. This avoids
+   confusing the model when it references a tool it discovered in an earlier turn.
+
+### Agent exemption
+
+Plan and task sub-agents do not use tool search. They operate on scoped tool
+sets (`AGENT_TOOLS` for plan agents, `TASK_AGENT_TOOLS` for task agents) with
+MCP tools merged in. Tool search is only active for the top-level session,
+where the model can interactively search for tools it needs.
 
 ---
 
 ## MCP Tools (External)
 
+> See also: [MCP Architecture diagram](diagrams/png/20-mcp-architecture.png)
+
 Turnstone supports the [Model Context Protocol](https://modelcontextprotocol.io/)
 (MCP) for connecting external tool servers — GitHub, databases, filesystems, or any
 MCP-compatible service.
@@ -418,13 +599,17 @@ MCP-compatible service.
 
 2. **Discovery**: At startup, `MCPClientManager` connects to each configured server
    (via stdio subprocess or HTTP), performs the MCP `initialize` handshake, and calls
-   `tools/list` to discover available tools.
+   `tools/list` to discover available tools. During the handshake, the manager checks
+   each server's capabilities for `tools.listChanged` support (push notifications).
 
 3. **Schema conversion**: Each MCP tool's `inputSchema` is converted to OpenAI
    function-calling format. The tool name is prefixed: `mcp__{server}__{tool}`.
 
-4. **Merging**: MCP tools are appended after the 14 built-in tools via
+4. **Merging**: MCP tools are appended after the 17 built-in tools via
    `merge_mcp_tools()`. Built-in tools appear first, giving them natural LLM priority.
+   When dynamic tool search is active, MCP tools are deferred rather than directly
+   visible -- the model discovers them via search as needed (see
+   [Dynamic Tool Search](#dynamic-tool-search) above).
 
 5. **Dispatch**: When the LLM calls an MCP tool, `_prepare_mcp_tool()` builds a
    generic approval preview and `_exec_mcp_tool()` calls `MCPClientManager.call_tool_sync()`,
@@ -497,3 +682,182 @@ MCP tools (3):
   mcp__github__create_issue  [MCP: github] Create a GitHub issue
   mcp__postgres__query       [MCP: postgres] Run a SQL query
 ```
+
+### Dynamic tool refresh
+
+MCP tool lists stay up-to-date without restart through three mechanisms:
+
+1. **Push notifications** -- MCP servers that declare `tools.listChanged: true` in
+   their capabilities send `notifications/tools/list_changed` when their tool list
+   changes. `MCPClientManager` registers a `message_handler` on each `ClientSession`
+   that triggers an immediate refresh for that server.
+
+2. **Periodic timer** -- Servers that do *not* support push notifications are polled
+   on a configurable interval (default 4 hours). The timer is staggered using a
+   launch-time seed (`monotonic_ns ^ pid`) so cluster nodes don't all hit MCP
+   servers simultaneously. Configure via `[mcp] refresh_interval` in `config.toml`
+   or `--mcp-refresh-interval SECONDS` on the CLI. Set to `0` to disable.
+
+3. **Manual** -- `/mcp refresh` re-fetches tools from all servers immediately.
+   `/mcp refresh <server>` targets a single server. If a server has disconnected,
+   manual refresh attempts reconnection.
+
+When tools change, `MCPClientManager` rebuilds its merged tool list using copy-on-write
+(new list/dict objects assigned atomically) and notifies all active `ChatSession`
+instances via registered listener callbacks. Each session rebuilds its `_tools`,
+`_task_tools`, `_agent_tools`, and reconstructs its `ToolSearchManager` (if active),
+preserving the set of previously expanded (discovered) tools.
+
+```toml
+[mcp]
+refresh_interval = 14400  # seconds (default 4h), 0 to disable
+```
+
+```
+/mcp refresh
+MCP refresh complete:
+  github: +1 added
+    + mcp__github__create_pr
+  postgres: no changes
+
+/mcp refresh github
+MCP refresh complete:
+  github: no changes
+```
+
+---
+
+## MCP Resources
+
+MCP servers can expose **resources** -- named data items (files, database rows,
+API responses) addressable by URI. turnstone discovers resources at startup and
+makes them available to the model via the `read_resource` built-in tool.
+
+### Discovery
+
+During the MCP `initialize` handshake, `MCPClientManager` checks each server's
+capabilities for the `resources` capability. For servers that declare it:
+
+1. `list_resources` fetches static resources (fixed URIs).
+2. `list_resource_templates` fetches URI templates (parameterized patterns like
+   `db://tables/{table}/rows/{id}`).
+
+Both are stored as `{uri, name, description, mimeType, server}` dicts and
+merged into a unified catalog.
+
+### Resource catalog in system message
+
+The first 50 resources are injected into the system message as an XML-delimited
+block so the model knows what URIs are available:
+
+```xml
+<mcp-resources>
+  file:///project/README.md  Project readme
+  db://users/schema  User table schema
+</mcp-resources>
+Use read_resource(uri='...') to access the resources listed above.
+```
+
+### read_resource tool
+
+| Parameter | Type   | Required | Description |
+|-----------|--------|----------|-------------|
+| `uri`     | string | yes      | The resource URI to read. |
+
+- **What it does**: Reads the resource from its MCP server via `MCPClientManager.read_resource_sync()`. Returns text content for text resources or base64-encoded data for binary resources. Output is truncated by the standard tool output limiter.
+- **Auto-approve**: No -- requires user confirmation (reads external data).
+- **Agent availability**: `agent` and `task_agent`.
+
+### Capability guards
+
+The `read_resource` tool schema is always loaded (it is a built-in JSON schema),
+but resource discovery only runs for servers that declare the `resources`
+capability. Servers without the capability contribute zero resources to the
+catalog.
+
+### Refresh
+
+Resource lists stay current through the same three-tier mechanism as tool lists:
+
+1. **Push** -- Servers declaring `resources.listChanged: true` send
+   `notifications/resources/list_changed`, triggering an immediate refresh.
+2. **Periodic** -- Servers without push are polled on the configured refresh
+   interval (default 4 hours, same timer as tools).
+3. **Manual** -- `/mcp refresh` re-fetches resources alongside tools.
+
+---
+
+## MCP Prompts
+
+MCP servers can also expose **prompts** -- reusable message templates with
+optional arguments. turnstone discovers prompts at startup for servers that
+declare the `prompts` capability.
+
+### Discovery
+
+Prompt discovery mirrors resource discovery: `list_prompts` is called during
+the `initialize` handshake. Each prompt is stored with its prefixed name
+(`mcp__{server}__{prompt}`), description, and argument schema.
+
+### use_prompt tool
+
+| Parameter   | Type   | Required | Description |
+|-------------|--------|----------|-------------|
+| `name`      | string | yes      | The prompt name (e.g. `mcp__server__prompt_name`). |
+| `arguments` | object | no       | Key-value argument pairs for the prompt. Values must be strings. |
+
+- **What it does**: Invokes an MCP prompt template by name via `MCPClientManager.get_prompt_sync()`, expanding it into messages. Returns the expanded prompt content formatted as `[role]: content` blocks joined with blank lines. The prompt catalog is listed in the system message so the model knows which prompts are available. Output is truncated by the standard tool output limiter.
+- **Auto-approve**: No -- requires user confirmation (invokes external prompt servers).
+- **Agent availability**: `agent` and `task_agent`.
+
+### Invocation
+
+`MCPClientManager.get_prompt_sync()` calls the server's `get_prompt` method
+with the provided arguments and returns the expanded messages. The `use_prompt`
+built-in tool exposes this to the model as a function call.
+
+### Governance Sync
+
+Discovered MCP prompts are automatically synced into the `prompt_templates`
+governance table as first-class governed templates:
+
+- **Origin tracking**: MCP-sourced templates have `origin="mcp"` and
+  `mcp_server` set to the server name. Manual templates have
+  `origin="manual"`.
+- **Read-only**: MCP-sourced templates are `readonly=True`. The admin API
+  returns 403 on update/delete attempts. The admin UI disables edit/delete
+  buttons and shows an origin badge.
+- **Precedence**: If a manual template and MCP prompt share the same name,
+  the manual template wins and the MCP prompt is skipped (with a log
+  warning).
+- **Lifecycle**: Templates are created on connect, updated on prompt list
+  refresh, and removed when the MCP server no longer exposes the prompt.
+  The sync runs automatically on connect, on `PromptListChangedNotification`,
+  and on manual `/mcp refresh`.
+- **Schema**: Migration 009 adds `origin`, `mcp_server`, and `readonly`
+  columns to the `prompt_templates` table.
+
+The `use_prompt` tool allows the model to invoke any discovered MCP prompt at
+runtime. A catalog of up to 30 prompts is injected into the system message
+inside `<mcp-prompts>` XML tags so the model can discover available prompts.
+
+---
+
+## MCP UI Visibility
+
+MCP server, resource, and prompt counts are surfaced across the UI:
+
+- **Server `/health` endpoint**: Returns `mcp.servers`, `mcp.resources`,
+  `mcp.prompts` when MCP is configured
+- **Server UI**: Magenta status badge in the header showing server count,
+  with resource/prompt counts in tooltip
+- **Console cluster status bar**: MCP metrics (servers/resources/prompts)
+  with magenta LED dot indicator, shown after a divider from workstream
+  metrics
+- **Console node detail**: Per-node MCP summary showing server, resource,
+  and prompt counts
+- **Console collector**: Aggregates MCP counts across all nodes in the
+  cluster overview
+
+MCP indicators use the `--magenta` design token for consistent theming
+across light and dark modes.