distillery-mcp 0.2.0

A configurable embedding system with DuckDB storage and MCP server integration

Distillery

Team Knowledge, Distilled
Capture, classify, connect, and surface team knowledge through conversational commands.

Documentation · Skills · Quick Start · Roadmap · Slides

---

What is Distillery?

Distillery is a team knowledge base accessed through Claude Code skills. It refines raw information from working sessions, meetings, bookmarks, and conversations into concentrated, searchable knowledge — stored as vector embeddings in DuckDB and retrieved through natural language. Runs locally over stdio or as a hosted HTTP service with GitHub OAuth for team access.

Distillery captures the highest-value transformation — from noise to signal — and makes it a tool the whole team can use.

Full documentation: norrietaylor.github.io/distillery

Skills

Distillery provides 10 Claude Code slash commands:

Skill	Purpose	Example
/distill	Capture session knowledge with dedup detection	/distill "We decided to use DuckDB for local storage"
/recall	Semantic search with provenance	/recall distributed caching strategies
/pour	Multi-entry synthesis with citations	/pour how does our auth system work?
/bookmark	Store URLs with auto-generated summaries	/bookmark https://example.com/article #caching
/minutes	Meeting notes with append updates	/minutes --update standup-2026-03-22
/classify	Classify entries and triage review queue	/classify --inbox
/watch	Manage monitored feed sources	/watch add github:duckdb/duckdb
/radar	Ambient feed digest with source suggestions	/radar --days 7
/tune	Adjust feed relevance thresholds	/tune relevance 0.4
/setup	Onboarding wizard for MCP connectivity and config	/setup

Quick Start

Step 1: Install the Plugin

claude plugin marketplace add norrietaylor/distillery
claude plugin install distillery

This installs all 10 skills. The plugin defaults to a hosted demo server — you can start using Distillery immediately.

Demo Server: distillery-mcp.fly.dev is for evaluation only. Do not store sensitive or confidential data.

Step 2: Switch to Local with uvx (Recommended)

For a private knowledge base, run the MCP server locally with uvx — no persistent install needed:

# Get a free API key from jina.ai, then:
export JINA_API_KEY=jina_...

Add to ~/.claude/settings.json (overrides the plugin's demo server):

{
  "mcpServers": {
    "distillery": {
      "command": "uvx",
      "args": ["distillery-mcp"],
      "env": {
        "JINA_API_KEY": "${JINA_API_KEY}"
      }
    }
  }
}

Restart Claude Code and run the onboarding wizard:

/setup

See the Local Setup Guide for full configuration options, or deploy your own instance for team use.

Development

uv pip install -e ".[dev]"
# or
pip install -e ".[dev]"
pytest                              # run tests
mypy --strict src/distillery/       # type check
ruff check src/ tests/              # lint

See Contributing for the full guide.

License

Apache 2.0 — see LICENSE for details.