john-slop/tests/fixtures/evaluation/README.md

834 B

Evaluation fixtures

These short fixtures were authored specifically for John Slop and are distributed under the repository's MIT license. They label observable rule signals only; no label claims that a human or an AI system authored the text.

This is a characterization corpus for regression testing, not a representative benchmark. Additions require review for provenance, licensing, expected rule IDs, and realistic negative examples.

Manifest schema 2 records which rule each clean case was reviewed as a realistic near miss for. Together the negative cases must cover every enabled shipped rule. signoff.toml binds the release decision to the SHA-256 digest of the exact heuristic catalog; changing defaults invalidates the automated gate until a reviewer reruns the corpus, assesses limitations, and explicitly renews it.