conductor-validator – skill Khác cho agent AI

Validate Conductor workflow artifacts

When to Use

Inspect a project that actually uses a conductor/ workflow directory after setup, before resuming a track, or when diagnosing inconsistent plans/status. This is distinct from the Rails Conductor desktop-app setup (conductor.json/script/server).

Read-only discovery

# Run from the explicitly selected project root; missing paths are findings, not permission to create them.
ls -ld conductor/ conductor/tracks/
ls conductor/index.md conductor/product.md conductor/tech-stack.md conductor/workflow.md conductor/tracks.md

Validation procedure

1. Read current project instructions and the actual configured Conductor workflow/version.

2. Compare the declared artifact list with the project’s real files; report missing, unreadable and extra artifacts separately.

3. Resolve local index/track links, track IDs and plan/status markers against their referenced content. Do not assume directory presence proves completeness.

4. Reconcile completed markers with recorded implementation/test evidence and the current revision; report stale or conflicting state without inventing a pass.

5. Output a concise list of file, observed inconsistency, expected contract and proposed correction. Apply fixes only when requested; preserve existing plans and dirty work.

No implementation-playbook file or runnable validator script is bundled. The procedure above is the complete local guide; optional project-specific automation must actually exist and be reviewed before execution.

Pattern Matching

Status markers in tracks.md:

- [ ] Track Name  # Not started
- [~] Track Name  # In progress
- [x] Track Name  # Complete

Task markers in plan.md:

- [ ] Task description  # Pending
- [~] Task description  # In progress
- [x] Task description  # Complete

Track ID pattern:

<type>_<name>_<YYYYMMDD>
Example: feature_user_auth_20250115

Limitations

• Use this skill only when the task clearly matches the scope described above.
• Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
• Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.