We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
docs: sandbox-rerun execution boundary and policy sharing (v1.14.0) Document review.sandbox_command: the argv-list form, the double opt-in (--sandbox-rerun plus a non-empty command), config set rejection, the 2-minute per-attempt bound, and the stderr notice. Test-name resolution is Go-only: non-Go files take the skip-and-warn path, so Python/JS/PHP/Java tests keep full static flaky detection but get no rerun confirmation. mcp and guard never execute the command at all. Also explain how to share .commitbrief/policy.yml past the wholesale .commitbrief/ gitignore, and why baseline.json deliberately stays ignored.
docs: sandbox-rerun flag + config for the flaky detector (v1.12.0)
Document architecture-aware review (archlint architecture.json bridge)
Document v1.7.0 glob filters, injection scan, secret patterns, alias
docs: flaky-test detector — config (review.flaky), --no-flaky flag, review command (ADR-0022)
Add `commitbrief summary` command for summarizing changes in plain language
Add default configuration for `commitbrief commit` command
Update Anthropic, Gemini, and OpenAI models to latest versions
Add `guard.token_preflight` and `--show-prompt` global flag
Add command.default configuration option for commitbrief command
Add cache stats and inspect subcommands, increase subcommands to four
Add OpenAI pricing override and deepseek provider in Config-files.md
Shrink `COMMITBRIEF.md` via `commitbrief compress` — [Compress command]
Add new documentation files