meta-skill-reviewer {Meta} (v1.0)

"Skills that create, review, search, and deploy other skills."

Purpose

Quality auditor for skills. Runs the 10/10 rubric (3 structural + 4 density + 3 safety) and 5 adversarial stress tests. [EXPLICIT]

When to use: /jm:review-skill "skill-name" or after creating a new skill.

Core Principles

1. Law of Rubric: Score = Structural (3pts) + Density (4pts) + Safety (3pts). Must be 10/10. [EXPLICIT]
2. Law of Adversarial: After rubric, run 5 stress tests (ambiguity, edge, contradiction, stress, security). [EXPLICIT]
3. Law of Non-Destruction: Read-only audit. Never modify the skill being reviewed. [EXPLICIT]

Core Process

Phase 1: Load skill SKILL.md

Phase 2: Score against rubric (structure, density, safety)

Phase 3: Run 5 adversarial tests. Report findings.

Validation Gate

• Operation completed successfully
• skills_index.json updated if needed
• context.json reflects current state
• No stack violations (R-002, R-003)

4. Self-Correction Triggers

[!WARNING] IF skills_index.json is stale THEN regenerate before searching. IF deploying a skill with status != production THEN WARN user.

Usage

Example invocations:

• "/meta-skill-reviewer" — Run the full meta skill reviewer workflow
• "meta skill reviewer on this project" — Apply to current context

Assumptions & Limits

• Assumes access to project artifacts (code, docs, configs) [EXPLICIT]
• Requires English-language output unless otherwise specified [EXPLICIT]
• Does not replace domain expert judgment for final decisions [EXPLICIT]

Edge Cases

Scenario	Handling
Empty or minimal input	Request clarification before proceeding
Conflicting requirements	Flag conflicts explicitly, propose resolution
Out-of-scope request	Redirect to appropriate skill or escalate