Documentation

Everything byre does is a generated file you can read; these pages are the reference for driving it. The deepest layer – architecture, skill authoring internals, the decision records – lives in the repo’s docs directory, and contributions are welcome: CONTRIBUTING.md.

User documentation

• Install -- go install, curl, Homebrew, or build from source
• Quickstart -- first run, the picker, and byre status
• Configuration -- the byre config editor -- widen or narrow the box in seconds
• Commands -- every byre command, generated from the binary
• Skills & templates -- your toolkit, in every folder -- and how to build and share it
• MCP & Claude Skills -- give your agent tools and skills, once, in every session
• Worktrees -- parallel agents, the git way
• Volumes & state -- what survives a rebuild, what's disposable, and which hammer resets what
• What's boxed, what isn't -- the honest contract, including what byre is not
• How do I…? -- the cookbook -- every recipe is a question, a tldr, and a shipped feature

Reference

• How it works -- config in, Dockerfile out, container up -- every step a file you can read
• Configuration reference -- the cascade, its merge rules, and the complete key vocabulary
• Packaging reference -- provenance, manifests, digest pinning, and the MCP / Claude Skills wiring contract
• Security model -- the threat model, the contract, and the sharp facts