| name | check-ruff-schema |
| description | Diagnose why henryiii/validate-pyproject-schema-store rejects a pyproject.toml because of unknown Ruff rule codes. |
check-ruff-schema
What this skill does
The validate-pyproject pre-commit / prek hook validates pyproject.toml against
JSON schemas bundled in henryiii/validate-pyproject-schema-store.
Its ruff.schema.json in resources/ enumerates known Ruff rule codes as an exhaustive enum,
so any code that Ruff has added but the schema store hasn't ingested yet causes
the whole [tool.ruff.lint] section to fail validation with an unhelpful error like:
tool.ruff.lint cannot be validated by any definition
This skill pinpoints which specific Ruff codes are the culprit by diffing the
project's extend-select / select / ignore / extend-ignore /
unfixable lists against the validate-pyproject-schema-store schema's enum.
When to run
Trigger whenever:
- The user says
prek run validate-pyproject
or pre-commit run validate-pyproject is failing on tool.ruff.lint.
- The user has just bumped Ruff (via
prek autoupdate, Renovate, manual
edit of .pre-commit-config.yaml) and wants to check compatibility.
- The user asks some version of "which of my Ruff rules aren't in the
schema yet?" or "why is validate-pyproject complaining?"
How to run
Invoke the bundled script against the repo's pyproject.toml (the default arg).
Note this won't rewrite the file, just print out a report
and gives a non-zero exit code if any codes are missing.
python scripts/check_ruff_codes.py
To see what Ruff currently supports, use ruff check --show-settings.
Codes vs. category prefixes
The schema enum accepts both full codes like E501 and bare category prefixes
like ANN, PTH, DTZ. The script treats both as valid, a "missing" report
means the string literally isn't in the validate-pyproject-schema-store schema's enum,
not that Ruff doesn't know it.
