Turnstone

Multi-node AI orchestration platform. Deploy tool-using AI agents across a cluster of servers, driven by message queues or interactive interfaces.

Named after the Ruddy Turnstone — a bird that flips rocks to expose what's hiding underneath.

What it does

Turnstone gives LLMs tools — shell, files, search, web, planning — and orchestrates multi-turn conversations where the model investigates, acts, and reports. It runs as:

• 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 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)

Quickstart

Interactive (terminal)

pip install turnstone
turnstone --base-url http://localhost:8000/v1

Interactive (browser)

turnstone-server --port 8080 --base-url http://localhost:8000/v1

Queue-driven (programmatic)

pip install turnstone[mq]
turnstone-bridge --server-url http://localhost:8080 --redis-host localhost

from turnstone.mq import TurnstoneClient

with TurnstoneClient() as client:
    # Generic — any available node picks it up
    result = client.send_and_wait("Analyze the error logs", auto_approve=True)
    print(result.content)

    # Directed — must run on a specific server
    result = client.send_and_wait(
        "Check disk I/O on this server",
        target_node="server-12",
        auto_approve=True,
    )

Cluster dashboard

pip install turnstone[console]
turnstone-console --redis-host localhost --port 8090

Then open http://localhost:8090 for the cluster-wide dashboard.

Docker

cp .env.example .env  # edit LLM_BASE_URL, OPENAI_API_KEY, etc.
docker compose up     # starts redis + server + bridge + console

Console dashboard at http://localhost:8090. See docs/docker.md for configuration, scaling, and profiles.

Simulator

Test the multi-node stack at scale without an LLM backend:

docker compose --profile sim up redis console sim

Or standalone:

pip install turnstone[sim]
turnstone-sim --nodes 100 --scenario steady --duration 60 --mps 10

See 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
│   ├── 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

Detailed UML diagrams are available in docs/diagrams/:

Diagram	Description
System Context	Top-level components and external dependencies
Package Structure	Python modules and dependency graph
Core Engine Classes	SessionUI protocol, ChatSession, WorkstreamManager
Conversation Turn	Full message lifecycle through the engine
Tool Pipeline	Three-phase prepare/approve/execute
MQ Protocol	9 inbound + 19 outbound message types
Message Routing	Multi-node routing scenarios
Redis Key Schema	All Redis keys, types, and TTLs
Workstream States	State machine transitions
Simulator	SimCluster, dispatchers, scenarios
Console Data Flow	Dashboard data collection threads
Deployment	Docker Compose service topology

Multi-node routing

Each Turnstone server runs a bridge process. Bridges share a Redis instance for coordination:

Redis Key	Purpose
turnstone:inbound	Shared work queue — generic tasks, any node
turnstone:inbound:{node_id}	Per-node queue — directed tasks
turnstone:ws:{ws_id}	Workstream ownership — auto-routes follow-ups
turnstone:node:{node_id}	Node heartbeat + metadata for discovery
turnstone:events:{ws_id}	Per-workstream event pub/sub
turnstone:events:global	Global event pub/sub
turnstone:events:cluster	Cluster-wide state changes (for turnstone-console)

Routing rules:

1. Message has target_node → routes to that node's queue
2. Message has ws_id → looks up owner, routes to owning node
3. Neither → shared queue, next available bridge picks it up

Bridges BLPOP from their per-node queue (priority) then the shared queue. Directed work always takes precedence.

Tools

14 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
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
task	Spawn autonomous sub-agent
plan	Explore codebase, write .plan.md
mcp__*	External tools from MCP servers

MCP Tool Servers

Turnstone supports the Model Context Protocol (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.

Configure via config.toml or --mcp-config:

[mcp.servers.github]
command = "npx"
args = ["-y", "@modelcontextprotocol/server-github"]

[mcp.servers.github.env]
GITHUB_TOKEN = "ghp_..."

Or use a standard MCP JSON config file:

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).

Multi-Model 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>.

[models.local]
base_url = "http://localhost:8000/v1"
model = "qwen3-32b"

[models.openai]
base_url = "https://api.openai.com/v1"
api_key = "sk-..."
model = "gpt-4o"
context_window = 128000

[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

Use /model to show available models, /model openai to switch. Workstreams created via the API accept an optional model parameter.

Configuration

All entry points read ~/.config/turnstone/config.toml. CLI flags override config values.

[api]
base_url = "http://localhost:8000/v1"
api_key = ""
tavily_key = ""

[model]
name = ""              # empty = auto-detect
temperature = 0.5
reasoning_effort = "medium"
default = "default"    # model alias for new workstreams
fallback = []          # ordered list of fallback model aliases
agent_model = ""       # model alias for plan/task sub-agents

[tools]
timeout = 30
skip_permissions = false

[server]
host = "0.0.0.0"
port = 8080

[redis]
host = "localhost"
port = 6379
password = ""

[bridge]
server_url = "http://localhost:8080"
node_id = ""           # empty = hostname_xxxx

[console]
host = "0.0.0.0"
port = 8090
url = "http://localhost:8090"  # used by CLI /cluster commands
poll_interval = 10

[mcp]
config_path = ""       # path to MCP JSON config file (alternative to TOML sections)

[mcp.servers.example]  # one section per MCP server
command = "npx"
args = ["-y", "@modelcontextprotocol/server-example"]
# type = "stdio"       # "stdio" (default) or "http"
# url = ""             # for HTTP transport

Precedence: CLI args > environment variables > config.toml > defaults.

Workstreams

Parallel independent conversations, each with its own session and state:

Symbol	State	Meaning
·	idle	Waiting for input
◌	thinking	Model is generating
▸	running	Tool execution in progress
◆	attention	Waiting for approval
✖	error	Something went wrong

Idle workstreams are automatically cleaned up after 2 hours (configurable). In multi-node deployments, workstream ownership is tracked in Redis — follow-up messages auto-route to the owning node.

Monitoring

/metrics endpoint exposes Prometheus-format metrics:

• turnstone_tokens_total{direction} — prompt/completion token counters
• turnstone_tool_calls_total{tool} — per-tool invocation counts
• turnstone_workstream_context_ratio{ws_id} — per-workstream context utilization
• turnstone_http_request_duration_seconds — request latency histogram
• turnstone_workstreams_by_state{state} — workstream state gauges

Per-workstream metrics are labeled by ws_id (bounded to 10 max workstreams).

Requirements

• Python 3.11+
• An OpenAI-compatible API endpoint (vLLM, NVIDIA NIM, llama.cpp, etc.)
• Redis (for message queue bridge — pip install turnstone[mq])

License

Business Source License 1.1 — free for all use except hosting as a managed service. Converts to Apache 2.0 on 2030-03-01.