Agent skill
codex-rules
Enforce repo execution discipline: keep PRs small; run migration-smoke when invariants change; never hand-edit vendored contracts.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/development/codex-rules
SKILL.md
Always
- Keep PRs small and auditable; avoid drive-by refactors.
- If invariants/plumbing are touched: run the canonical proof command listed in EVALSETS.yaml (currently:
npx nx run evalsets:migration-smoke). - Never hand-edit vendored contracts; use sync scripts + drift gates.
Claude lane (enforced triggers)
If any of the following are true, you MUST propose a Claude Code run before proceeding:
- Repo-wide enumeration / dependency mapping across many files.
- Independent audit/review requested (migration correctness, invariants, compliance).
- Large-context synthesis from long documents.
When proposing Claude, provide all of:
- Claude prompt (ready to copy/paste)
- Artifact path to save output into (default:
docs/reviews/YYYYMMDD-HHMM_<slug>.md) - Acceptance criteria (what the artifact must contain)
Do not proceed until the artifact exists, unless the user explicitly declines Claude. If declined, proceed in Codex but label increased risk.
Claude Code (optional specialist)
If the user asks for large mechanical edits/refactors, or if the change would touch many files, propose a Claude Code handoff.
When proposing Claude, output a paste-ready “Claude Work Order” with:
- Goal
- Allowed paths
- Hard constraints (no churn, no invariant changes, etc.)
- Verification commands
- Expected deliverable summary Do not proceed with Claude unless the user explicitly says “yes”.
Claude Code (optional specialist)
Follow the AGENTS.md handoff protocol. Codex must provide a paste-ready work order (goal, allowed paths, constraints, verification, expected output), then re-verify and update the worklog after Claude returns results.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?