.coderabbit.yaml

# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json
early_access: true
reviews:
  profile: assertive
  path_instructions:
    - path: "**/vendor_files/**"
      instructions: "These files came from a vendor and we're not allowed to change them. Refer to it if you need to understand how the main code interacts with it, but do not make comments about it."
    - path: "**/*.py"
      instructions: "Check the `ruff.toml` and `ruff-test.toml` for linting rules we've explicitly disabled and don't suggest changes to please conventions we've disabled. Do not express concerns about ruff rules; a pre-commit hook already runs a ruff check. Do not warn about unnecessary super().__init__() calls; pyright prefers those to be present. Do not warn about missing type hints; a pre-commit hook already checks for that."
  tools:
    eslint: # when the code contains typescript, eslint will be run by pre-commit, and coderabbit often generates false positives
      enabled: false
    ruff: # when the code contains python, ruff will be run by pre-commit, and coderabbit often generates false positives
      enabled: false
    pylint: # when the code contains python, pylint will be run by pre-commit, and coderabbit often generates false positives
      enabled: false
    flake8: # we use ruff instead (when we use Python)
      enabled: false
  poem: false
  in_progress_fortune: false
  # the commit status is driven by our repository config and required checks, we don't want CodeRabbit messing with it
  commit_status: false
  auto_review:
    # a main purpose of opening a draft PR might be to get CodeRabbit feedback early
    drafts: true
  finishing_touches:
    docstrings:
      enabled: false # if we wanted AI to generate docstrings, it would be via CLI, not in the GitHub interface
    unit_tests:
      enabled: false # Quis custodiet ipsos custodes? not something we want AI doing, especially not via the GitHub interface