chore: add a CodeRabbit config on top of the organization settings
Merged
BO
16d ago+51+51+01234567890123456789-0-0-0123456789
Adds .coderabbit.yaml. Until now CodeRabbit used only the organization settings from its web UI.
inheritance: true, so every key the file leaves out keeps its organization value. Without it, a repository file replaces the organization settings and missing keys fall back to CodeRabbit's defaults.- The summary comment drops the effort estimate, sequence diagrams, suggested reviewers and labels, and the waiting fortune. The docstring coverage check is off.
- One instruction for all paths: flag only problems this PR introduces, and check a finding against the code before posting it. Golden
.dumpfiles get a second one: every changed line has to follow from the PR's extraction, resolution or synthesis change. Fixtures stay in review because the dump diff is the review artifact (AGENTS.md). - Two custom pre-merge checks, both warnings: a new lint, type-check or compiler suppression needs a comment saying why, and a user-visible change (CLI, MCP tool, language, framework, agent target, config key) needs
README.mdandsite/src/content/docs/updates, following the AGENTS.md rule. - New learnings wait 7 days for approval before they apply.
AGENTS.mdandCLAUDE.mdare already read as code guidelines by default.
README rows checked: none apply. This changes review tooling only, not a feature, command, language or measured number.
Check: the file validates against https://coderabbit.ai/integrations/schema.v2.json, and every key exists in it. CodeRabbit reads the config from the PR branch, so its review of this PR runs with the new settings.