Audit a Claude Code plugin for quality and correctness. Use when the user wants to review, validate, or get improvement suggestions for a plugin's skills and agents. Evaluates structure, purpose correctness (Skill vs Agent), invocation design, content quality, and cross-plugin coherence. Produces a structured report with a scorecard.
How this skill is triggered — by the user, by Claude, or both
Slash command
/advanced-plugin-creator:validate-pluginThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Audit the plugin at `$ARGUMENTS` (or ask the user to provide the path if not given).
Audit the plugin at $ARGUMENTS (or ask the user to provide the path if not given).
Read every skills/*/SKILL.md and every agents/*.md file in the plugin directory. Then produce a structured review report following the parts below.
For each skill in skills/*/SKILL.md, evaluate:
skills/<name>/SKILL.md)?name follow the rules: max 64 chars, lowercase letters/numbers/hyphens only, no XML tags, no reserved words?description present, non-empty, under 1024 chars, and free of XML tags?description written so Claude can semantically match it to user requests? Does it contain natural trigger phrases?disable-model-invocation: true is set, is manual-only invocation appropriate for this use case?agent: is specified, is forking to a subagent appropriate here?Provide up to 3 concrete improvement suggestions per skill.
For each agent in agents/*.md, evaluate:
.md file (not a directory) inside agents/?name and description?description clear about when Claude should invoke this agent (not just what it does)?allowed-tools) specified appropriately — not too permissive, not unnecessarily limited?Provide up to 3 concrete improvement suggestions per agent.
plugin-name:skill-name pattern if needed? Any naming conflicts?Produce the report in this structure:
## Plugin Review: <plugin name>
### Skills
#### <skill-name>
- Structure: ✅ / ⚠️ / ❌ [brief note]
- Purpose: ✅ correct Skill usage / ⚠️ should be an Agent / ❌ [issue]
- Invocation: ✅ / ⚠️ / ❌ [brief note]
- Content quality: ✅ / ⚠️ / ❌ [brief note]
- Suggestions:
1. ...
2. ...
### Agents
#### <agent-name>
- Structure: ✅ / ⚠️ / ❌ [brief note]
- Purpose: ✅ correct Agent usage / ⚠️ should be a Skill / ❌ [issue]
- System prompt quality: ✅ / ⚠️ / ❌ [brief note]
- Tool permissions: ✅ / ⚠️ / ❌ [brief note]
- Suggestions:
1. ...
2. ...
### Cross-Plugin Analysis
- Role confusion: ...
- Overlap: ...
- Missing pieces: ...
- Namespace: ...
### Summary scorecard
| Component | Type | Purpose ✅/⚠️/❌ | Quality ✅/⚠️/❌ | Priority fix |
|-----------|------|----------------|----------------|--------------|
| ... | Skill/Agent | ... | ... | ... |
Be specific and actionable. Reference exact file paths, frontmatter fields, and line-level issues where possible.
npx claudepluginhub tymoj/marketplace --plugin advanced-plugin-creatorAudits agents and skills across the plugin ecosystem for quality, consistency, and correct invocation configuration. Use after creating or modifying multiple skills.
Validates Claude Code plugin structure and manifest, launches agent to fix issues with user confirmation, increments version, re-validates, and reports. For plugin quality optimization.
Audits Claude Code plugins for structure validation, frontmatter quality, deprecations, feature adoption, security patterns, and documentation. Ensures changelog compatibility and best practices for releases.