Get ThreadWeave running in under two minutes.
One command. Works on Linux, macOS, and Windows.
git clone https://github.com/PowerLooming/ThreadWeave.git cd ThreadWeave docker compose up
API available at http://localhost:8000. API docs at /docs.
Optionally run with a local LLM for smarter detection:
docker compose --profile llm up
Pulls Ollama + llama3.1:8b automatically on first start. Data persists in Docker volumes across restarts.
For development or when Docker isn't available. Requires Python 3.11+ and uv.
git clone https://github.com/PowerLooming/ThreadWeave.git cd ThreadWeave bash setup.sh
Or manually:
uv venv --python 3.11 .venv source .venv/Scripts/activate # Windows uv pip install -e ".[dev]"
Start the server:
uv run python -m threadweave.cli serve
ThreadWeave ships with:
• API server — FastAPI with auto-generated docs
• CLI — detect, search, save, serve commands
• Detection engine — Regex + optional LLM classifier
• M365 connectors — Exchange Online, SharePoint, Microsoft Graph
• Google Workspace connectors — Gmail, Chat, Drive
• Confidentiality engine — Seven sensitivity levels with access enforcement
• Profiling — Latency percentiles, Prometheus export
• 232 tests — Mock-based, no external APIs called
After installation, follow the Setup Guide to connect your Microsoft 365 or Google Workspace account. Read the Governance FAQ for access control, confidentiality, and GDPR details. Explore the configuration guide for all thresholds and environment variables.