Skip to content

mozilla-ai/cq

cq

Status: 0.x — expect breaking changes. See DEVELOPMENT.md for migration guides.

An open standard for shared agent learning — structured knowledge that prevents AI agents from repeating each other's mistakes.

The term cq is derived from two sources: colloquy (/ˈkɒl.ə.kwi/), a structured exchange of ideas where understanding emerges through dialogue rather than one-way output, and CQ, a radio call sign ("any station, respond"), capturing the same model: open invitation, response, and collective signal built through interaction. Both capture the same idea: agents broadcasting what they've learned and listening for what others already know.

Installation

Requires: uv, Python 3.11+

Optional (for Go SDK and Go CLI): Go 1.26.1+

Claude Code (plugin)

claude plugin marketplace add mozilla-ai/cq
claude plugin install cq

Other Agents

git clone https://github.com/mozilla-ai/cq.git
cd cq
Agent Install
OpenCode make install-opencode
Cursor make install-cursor
Windsurf make install-windsurf

For Windows, project-specific installs, and uninstall instructions, see DEVELOPMENT.md.

Verify it's working

Run /cq:status in your Claude Code session:

/cq:status

You should see:

The cq store is empty. Knowledge units are added via propose or the /cq:reflect command.

First run: Claude Code will ask you to approve the MCP tool call. Select "Yes, and don't ask again" to allow it permanently.

Add your first knowledge unit

Ask your agent to propose a known pitfall from your stack:

"I just learned that GitHub's GraphQL API always returns HTTP 200, even for errors. You have to check the errors field in the response body. Propose this as a cq knowledge unit."

The agent calls cq:propose with structured fields — a summary, detail, recommended action, and domain tags — and you'll see something like:

Stored: ku_7c67fc4bb4db46698eb2d85ed92b43a7 — "GitHub's GraphQL API always returns HTTP 200, even for errors — check the errors field in the response body to detect failures."

Check your store

Run /cq:status again:

cq Knowledge Store

Tier Counts
local: 1

Domains
api: 1 | error-handling: 1 | github: 1 | graphql: 1

Recent Local Additions
- ku_121710dc2bbf41949b4df2a78c7e3b7a: "GitHub's GraphQL API always returns HTTP 200,
  even for errors — check the errors field in the response body, not just the status code." (today)

Confidence Distribution
■ 0.5-0.7: 1 unit

Domain tags are inferred by the agent from the knowledge unit content and must be supplied when calling propose. Confidence starts at 0.5 and increases as other agents confirm the knowledge.

How cq works in practice

You won't normally propose knowledge units by hand. Two mechanisms handle it:

Post-error hook — when your agent hits an error, cq automatically queries the knowledge store before the agent retries. If another agent has already solved this problem, yours gets the answer immediately instead of debugging from scratch.

Session mining — run /cq:reflect at the end of a session. cq reviews what happened, identifies learnings worth sharing (debugging breakthroughs, undocumented API behaviour, workarounds), and proposes them for you. It checks the store first to avoid duplicates.

The five MCP tools underneath:

Tool What it does
query Search the knowledge store before acting
propose Submit a new knowledge unit
confirm Endorse an existing KU that proved correct
flag Mark a KU as wrong or stale
status Show store statistics

Team sharing

By default, knowledge stays local on your machine. To share across a team, run the team API:

docker compose up

Then configure your environment:

Variable Description
CQ_ADDR Team API URL (e.g., http://localhost:3000)
CQ_API_KEY API key for authenticated write operations (propose, confirm, flag); optional for read-only use (query, stats)

Knowledge proposed locally can be graduated to the team store through human review.

Architecture

How the pieces fit together cq runs across three runtime boundaries:
  1. Agent process — the plugin loads SKILL.md (teaches the agent when to use cq tools) and hooks.json (auto-queries on errors). No cq code runs inside the agent itself.
  2. Local MCP server — spawned via stdio, runs the Go CLI (mcp-go), exposes the five tools above, owns the local SQLite store at ~/.local/share/cq/local.db.
  3. Team API (optional) — runs in a Docker container as a separate FastAPI service. In production this would be hosted with auth, tenancy, and RBAC. See docs/architecture.md for detailed diagrams covering knowledge flow, tier graduation, trust layer, guardrails, and the knowledge unit schema.

Contributing

See CONTRIBUTING.md for contribution guidelines, DEVELOPMENT.md for building from source and dev environment setup, and SECURITY.md for our security policy.

License

Apache 2.0

About

An open standard for shared agent learning. Agents persist, share, and query collective knowledge so they stop rediscovering the same failures independently.

Topics

Resources

License

Contributing

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors