mirror of
https://github.com/turnstonelabs/turnstone.git
synced 2026-08-12 23:12:23 -06:00
Patrick Buckley
42a87d0e1e
CI's postgres-backend run failed 11 of the new notify tests from #505. Three independent issues: 1. Migration 053's ``services_notify`` trigger lives only in the alembic chain, but the test fixture in conftest.py calls ``init_storage(..., run_migrations=False)`` for speed. That path skips migrations and relies on ``metadata.create_all`` for the table tree. Previous alembic-only DDL (migrations 041 / 048 ``CREATE INDEX CONCURRENTLY`` on workstreams) is performance-only, so tests never depended on it. 053's trigger is the first behaviorally-required alembic-only DDL in the project — without it ``register_service`` doesn't fire NOTIFY and the trigger-filter tests time out. Fix: declare the trigger function + trigger in ``_schema.py`` and attach them via ``sa.event.listen(services, "after_create", ...)`` DDL events, gated on ``dialect == "postgresql"``. The same SQL constants are imported by migration 053 so there's a single source of truth. Test fixture stays unchanged — ``create_all`` now installs the trigger on fresh PG test DBs. Migration covers the upgrade-on-existing-DB path; the two are mutually exclusive given ``create_tables = not run_migrations`` in ``init_storage``. 2. NotifyDispatcher tests fired ``storage.notify(...)`` immediately after ``d.start()`` and hit a race: the listener thread is concurrently calling ``psycopg.connect(listen_url)`` + ``LISTEN <channel>`` over the network, so the notify can land before any session is listening on the channel and PG drops it (pg_notify only routes to sessions LISTEN'ing at COMMIT time). Fix: dispatcher gains a ``_listener_ready: threading.Event`` set inside ``_listener_loop`` after each successful ``storage.listen`` open and cleared on disconnect, plus a public ``wait_until_ready(timeout)`` method. Tests use a new ``_start_ready(d)`` helper that calls ``start()`` + asserts ready. Production callers don't need this (real reactive traffic arrives well after startup), but it's the right primitive for any future "start dispatcher, immediately send" call site too. 3. ``TestSqliteNotify`` is misnamed — its tests run against whichever backend the ``storage`` fixture provides (PG by default in CI). Two of its assertions were SQLite-specific: ``assert got.pid == 0`` only holds for the synthetic in-process path (PG carries real backend PIDs), and ``test_synthetic_sweep_emits_after_interval`` is fundamentally SQLite-only (no sweep on the PG path). Fix: drop the pid assertion (channel + payload are the backend-agnostic invariants), add an ``_is_sqlite`` fixture mirror of ``_is_postgres``, and gate the sweep test on it. The sweep test also moves from monkey-patching ``stream._sweep_interval`` to passing the ``sweep_interval`` kwarg that ``SQLiteBackend.listen`` now accepts (from the earlier Copilot review fix). Validated locally against a fresh ``turnstone_test`` PG DB: 263 storage + console + notify tests pass on PG, 257 on SQLite, mypy + ruff clean.
Turnstone
Multi-node AI orchestration platform. Deploy tool-using AI agents across a cluster of servers with direct HTTP routing, interactive interfaces, and enterprise governance.
Named after the Ruddy Turnstone (Arenaria interpres) — a shorebird that flips stones to discover what's hiding underneath.
Release Tracks
| Track | Install | Docker | Description |
|---|---|---|---|
| Stable | pip install turnstone |
ghcr.io/turnstonelabs/turnstone:stable |
Production-grade. Bugfixes only. |
| Experimental | pip install turnstone --pre |
ghcr.io/turnstonelabs/turnstone:experimental |
New features. May have rough edges. |
See docs/releasing.md for the full release process.
What it does
Turnstone gives LLMs tools — shell, files, search, web, planning — and orchestrates multi-turn conversations where the model investigates, acts, and reports.
- Interactive sessions — terminal CLI or browser UI with parallel workstreams
- Cluster dashboard — real-time view of all nodes and workstreams with console routing proxy
- Intent validation — LLM judge evaluates every tool call with risk assessments and evidence
- Governance — RBAC, OIDC SSO, tool policies, skills, usage tracking, audit logs
- Multi-provider — OpenAI-compatible APIs (vLLM, llama.cpp, NIM), Anthropic Messages API, and Google Gemini
- MCP support — external tool servers with native deferred loading (Anthropic/OpenAI) or BM25 fallback
Quickstart
pip install turnstone
# Terminal REPL
turnstone --base-url http://localhost:8000/v1
# Browser UI
turnstone-server --port 8080 --base-url http://localhost:8000/v1
# Cluster dashboard
pip install turnstone[console]
turnstone-console --port 8090
For PostgreSQL (recommended for production):
pip install turnstone[postgres]
export TURNSTONE_DB_BACKEND=postgresql
export TURNSTONE_DB_URL="postgresql+psycopg://user:pass@localhost:5432/turnstone"
turnstone-server --port 8080 --base-url http://localhost:8000/v1
Docker
cp .env.example .env # edit LLM_BASE_URL, OPENAI_API_KEY, etc.
docker compose --profile production up
See QUICKSTART.md for the bootstrap wizard and docs/docker.md for Docker configuration and profiles.
Programmatic (SDK)
from turnstone.sdk import TurnstoneServer
with TurnstoneServer("http://localhost:8080", token="tok_xxx") as client:
ws = client.create_workstream(name="demo")
result = client.send_and_wait("Analyze the error logs", ws.ws_id, auto_approve=True)
print(result.content)
Tools
Built-in tools for shell, files, search, web, memory, notifications, and autonomous sub-agents — plus external tools via MCP with native deferred loading. See docs/tools.md for the full reference and docs/mcp-registry.md for MCP configuration.
Architecture
Single-node: Client → Server (direct HTTP + SSE). No external dependencies beyond the database.
Multi-node: Client → Console (rendezvous routing proxy) → Server nodes. The console picks the target node for each workstream via rendezvous (HRW) hashing over the live service registry — pure function of (ws_id, live_nodes), no stored bucket state, deterministic across readers. A node join or drop only re-routes the keys that score highest on the affected node.
| Component | Purpose |
|---|---|
turnstone |
Terminal CLI (REPL) |
turnstone-server |
Web UI + REST API + SSE events |
turnstone-console |
Cluster dashboard + routing proxy + admin panel |
turnstone-channel |
Channel gateway (Discord and Slack adapters) |
turnstone-admin |
User/token management CLI |
turnstone-eval |
Eval harness for prompt/tool optimization |
turnstone-bootstrap |
LLM-guided setup wizard |
Diagrams
UML diagrams in docs/diagrams/:
| Diagram | Description |
|---|---|
| System Context | Components and external dependencies |
| Package Structure | Python modules and dependency graph |
| Core Engine | SessionUI, ChatSession, LLMProvider |
| Conversation Turn | Message lifecycle through the engine |
| Tool Pipeline | Prepare / approve / execute |
| Workstream States | State machine transitions |
| Console Data Flow | Dashboard data collection |
| Deployment | Docker Compose topology |
| Auth | JWT, scopes, login flows |
| Channels | Discord / Slack adapters + routing |
| Judge | Intent validation pipeline |
| OIDC | SSO authorization code flow |
Documentation
| Topic | Link |
|---|---|
| Configuration reference | docs/settings.md |
| API reference | docs/api-reference.md |
| Docker deployment | docs/docker.md |
| Intent validation (judge) | docs/judge.md |
| Governance & RBAC | docs/governance.md |
| OIDC SSO | docs/oidc.md |
| TLS / mTLS | docs/tls.md |
| Channel integrations | docs/channels.md |
| Console dashboard | docs/console.md |
| Eval harness | docs/eval.md |
| Tools reference | docs/tools.md |
| MCP integration | docs/mcp-registry.md |
Requirements
- Python 3.11+
- An OpenAI-compatible API endpoint, Anthropic API key, or Google Gemini API key
- Optional: PostgreSQL (
pip install turnstone[postgres]), Anthropic (pip install turnstone[anthropic]) - Git LFS for cloning (diagram PNGs)
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.