Documentation
Briar — the operator's manual.
Quickstart
Three commands get you from a fresh repo to an open PR. Each one is documented in depth on its own page — these are the shortest path.
1. Install
$ pip install briar-cli$ briar version
2. Authenticate the providers you'll use
$ briar auth login github-pat --company acme$ briar auth login jira-token --company acme# LLM key is read from the environment, not via auth login:$ export ANTHROPIC_API_KEY=sk-ant-...
Tokens are written to ~/.config/briar/secrets.envby default. Audit coverage with briar secrets doctor (it walks the runbook YAMLs under ./examples).
3. Build a plan and run the loop
$ briar plan build jira:KAN --name q3 --company acme --llm anthropic$ briar plan run q3 --company acme --llm anthropic \--owner acme --repo widgets
The selector picks the next card, the engineer agent opens the PR, the knowledge file learns what changed. Loop continues until the board is done.
Want a guided walkthrough?
How these docs are organised
- Get started — install, mental model, configuration. Read top-to-bottom on your first sitting.
- CLI reference — one page per command group. Every flag, default, and choice listed; examples for the common cases.
- Reference — plugin registries (LLMs, providers, trackers, ...), runbook YAML schema, and a recipes page for full workflows.
Next steps
Install & first run →
pip install briar-cli, point it at a tracker board, and watch the first plan build.
Core concepts →
Knowledge files, runbooks, plans, agents, journals — the five primitives Briar wires together.
Configuration →
Environment variables, credential stores, BRIAR_* settings, and the secrets cascade.
CLI reference →
Every command, every flag, every default — generated against the live --help text.
Source of truth
briar <cmd> --help output. Spot a drift? Open an issue.