AGENTS.md Lint (Schliff)
Version updated for https://github.com/Zandereins/schliff to version v8.7.0.
- This action is used across all versions by 2 repositories.
Action Type
This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary
This GitHub Action automates the scoring of AGENTS.md files using the Schliff tool. It ensures that AI instruction files remain consistent across different systems by providing deterministic quality scores based on a versioned rubric. The action helps detect and prevent issues in instruction files that could degrade LLMs, ensuring better reliability in AI development workflows.
What’s Changed
Pre-launch audit batch A + C.
Changed
structureis now reproducible — a pure function of the file’s bytes (#10). It previously depended on the file’s on-disk neighbourhood (a siblingreferences/directory + whether declared references resolved on disk), so the same file scored ~15 points higher in its real directory than in isolation — the public badge disagreed with the author’s own CLI, worst on AGENTS.md. Now progressive disclosure is credited from content disclosure-links (not an on-disk dir), references from the declaration (with a traversal/stuffing guard), and dangling detection is a non-scoring lint issue emitted only from a provable location. Also corrects a long-standing under-crediting of AGENTS.md/CLAUDE.md/.cursorrules(always temp-scored). Scores for files that link detail move upward and now match in-repo and in isolation. Golden rebaselined with documented values; field-validated over 115 real skills; a new isolation-equivalence test pins location-independence.
Fixed
- The terminal no longer silently drops score warnings — the calibrated-weights “not comparable” notice and “no weighted dimensions” warning now render (#22).
compareno longer leaks the-1sentinel for unmeasured dimensions (#21).
Note for CI users: this is a scoring-model change; scores shift upward for
files that link detail. Pin schliff-version: '8.7.0' if you gate on an exact number.