skills/boshu2__agentops
1
0
Fork 0
mirror of https://github.com/boshu2/agentops.git synced 2026-09-14 15:08:13 +08:00
Code Issues Packages Projects Releases Wiki Activity
Files
main
boshu2__agentops/scripts/lint/README-codex-residual-policy.md
T
Boden Fuller 9ea9fbc957 feat: add codex residual marker linting gate 
- codex-residual-allowlist.txt with policy docs
- validate-codex-runtime-sections.sh CI gate
- README-codex-residual-policy.md authoring guide
2026-02-26 11:02:06 -05:00

1.2 KiB
Raw Permalink Blame History

Codex residual marker allowlist policy

This policy governs scripts/lint/codex-residual-allowlist.txt, the canonical machine-readable allowlist for residual mixed-runtime markers in skills-codex/**/SKILL.md.

Purpose

skills-codex is Codex-first, but a small set of Claude markers is intentionally retained for mixed-runtime flows (--mixed) and runtime-native fallback documentation. The allowlist defines those exceptions explicitly so lint can fail on accidental runtime drift.

Authoring rules

  1. One POSIX ERE pattern per line in the allowlist file.
  2. Keep patterns narrow and stable. Prefer exact phrases, backend IDs, reference filenames, and primitive names.
  3. Do not use broad wildcards or generic vendor tokens (Claude, claude, .*, .*claude.*).
  4. Every new entry must map to an intentional mixed-runtime contract in skills-codex/**/SKILL.md.
  5. Remove entries once no longer referenced.

Review checklist for allowlist changes

  1. Is this marker required for mixed-runtime behavior?
  2. Is the pattern as specific as possible?
  3. Could this be expressed by an existing allowlisted marker?
  4. Will this hide unintended Codex-to-Claude regressions?
Reference in New Issue View Git Blame Copy Permalink