Stats
Actions
Tags
From plugin-auditor
Audits plugin suites using subagents to analyze usability, coverage gaps, duplications, and workflow optimizations. Generates reports with user-approved todo conversion. Useful for plugin ecosystem improvements.
How this skill is triggered — by the user, by Claude, or both
Slash command
/plugin-auditor:suite-auditThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Meta-level audit skill that analyzes any plugin suite using specialized subagents to identify usability improvements, coverage gaps, duplication, and workflow optimizations.
Meta-level audit skill that analyzes any plugin suite using specialized subagents to identify usability improvements, coverage gaps, duplication, and workflow optimizations.
Systematically evaluate a plugin ecosystem to:
This skill orchestrates four specialized analysis passes using subagents, then synthesizes findings into a prioritized report.
Goal: Evaluate each agent/skill for user experience quality.
| Criterion | Question | Weight |
|---|---|---|
| Discoverability | Can users find the right agent easily? | 20% |
| Clarity | Are instructions clear and unambiguous? | 20% |
| Completeness | Does the agent deliver on its promise? | 20% |
| Flexibility | Can it handle edge cases and variations? | 15% |
| Integration | Does it work well with other agents? | 15% |
| Output Quality | Are outputs immediately usable? | 10% |
Analyze [agent-name] for usability issues:
1. Read the agent definition thoroughly
2. Evaluate against these criteria:
- Discoverability: Clear description with good trigger words?
- Clarity: Unambiguous instructions?
- Completeness: Covers expected use cases?
- Flexibility: Handles variations?
- Integration: Connects to other agents?
- Output Quality: Well-defined formats?
3. Score each criterion 1-5
4. Identify specific improvement opportunities
5. Suggest concrete fixes
Output format:
## Usability Analysis: [Agent Name]
**Overall Score**: X/30
**Critical Issues**: [list]
**Improvements**: [list with specific recommendations]
Goal: Identify missing agents that would strengthen the suite.
A gap is worth filling if:
Analyze [plugin-name] for coverage gaps:
1. Determine the plugin's domain and purpose
2. Map the complete workflow for this domain
3. Identify which workflow stages have agent coverage
4. Find stages without dedicated support
5. Evaluate gap significance:
- How often do users need this?
- Is it complex enough for an agent?
- Could existing agents cover it with modifications?
6. Recommend new agents with specifications
Output format:
## Gap Analysis: [Plugin Name]
**Domain**: [what this plugin covers]
**Coverage Score**: X% of workflow covered
**Workflow Map**: [stages with coverage indicators]
**Critical Gaps**: [list with justification]
**Recommended New Agents**: [name, purpose, why needed]
Goal: Find overlapping functionality that can be consolidated.
| Type | Description | Action |
|---|---|---|
| Full Overlap | Two agents do same thing | Merge or deprecate one |
| Partial Overlap | Significant shared functionality | Extract common patterns |
| Conceptual Overlap | Similar approaches in different contexts | Document boundaries |
| Output Overlap | Produce similar artifacts | Standardize formats |
Scan for:
Analyze [plugin-name] for duplication:
1. Read all agent definitions in the plugin
2. Create a feature matrix showing what each agent provides
3. Identify overlapping functionality:
- Exact duplicates
- Similar but not identical features
- Overlapping scopes
4. Recommend consolidation strategies:
- Merge agents
- Extract shared components
- Clarify boundaries
5. Estimate effort to consolidate
Output format:
## Duplication Analysis: [Plugin Name]
**Redundancy Level**: [Low/Medium/High]
**Full Overlaps**: [specific examples with quotes]
**Partial Overlaps**: [specific examples]
**Consolidation Recommendations**: [actions with rationale]
Goal: Improve how agents work together.
Evaluate:
WORKFLOW CHAINS
===============
Chain Name:
agent-1 → agent-2 → agent-3
Evaluation:
- Handoffs clear: [Yes/No]
- Output compatibility: [Yes/No]
- Documentation: [Yes/No]
Analyze [plugin-name] agent workflow integration:
1. Map typical user journeys through the plugin
2. Identify how agents currently reference each other
3. Find workflow friction points:
- Unclear handoffs
- Missing integration
- Redundant steps
4. Recommend optimizations:
- New integration points
- Workflow templates
- Automation opportunities
Output format:
## Workflow Analysis: [Plugin Name]
**Integration Score**: X/10
**Workflow Chains Identified**: [list]
**Friction Points**: [specific issues]
**Optimization Recommendations**: [actions]
1. Read plugin.json for metadata and component list
2. Glob all agents: [plugin-path]/agents/*.md
3. Glob all skills: [plugin-path]/skills/*/SKILL.md
4. Glob all commands: [plugin-path]/commands/*.md
5. Build inventory of all components
Launch four analysis passes using the Task tool:
Combine analyses into unified audit report:
# Plugin Suite Audit Report: [Plugin Name]
**Date**: [Date]
**Version Audited**: [version from plugin.json]
**Components Analyzed**: X agents, Y skills, Z commands
## Executive Summary
**Overall Health Score**: X/100
| Category | Score | Issues |
|----------|-------|--------|
| Usability | X/25 | X critical, X major |
| Coverage | X/25 | X gaps identified |
| Duplication | X/25 | X redundancies found |
| Workflow | X/25 | X friction points |
**Key Findings**:
1. [Most critical finding]
2. [Second finding]
3. [Third finding]
---
## Detailed Findings
### Usability Issues
[From Phase 1]
### Coverage Gaps
[From Phase 2]
### Duplication Concerns
[From Phase 3]
### Workflow Friction
[From Phase 4]
---
## Prioritized Recommendations
### Critical (Do First)
| # | Recommendation | Category | Effort | Impact |
|---|----------------|----------|--------|--------|
| 1 | [Action] | [Category] | [Low/Med/High] | [High] |
### Important (Do Soon)
[Table continues]
### Nice to Have (Consider)
[Table continues]
---
## Suggested New Agents
| Agent Name | Purpose | Gap Filled | Priority |
|------------|---------|------------|----------|
| [name] | [what it does] | [workflow gap] | [High/Med/Low] |
---
## Consolidation Targets
| Source A | Source B | Overlap | Recommendation |
|----------|----------|---------|----------------|
| [agent] | [agent] | [what overlaps] | [merge/extract/clarify] |
After presenting the report, use AskUserQuestion:
Which findings should I convert to actionable todos?
Options:
1. All recommendations (Critical + Important + Nice to Have)
2. Critical and Important only
3. Critical only
4. Let me select specific items
5. Report only, no todos
Upon approval, create todos with:
[Usability], [Gap], [Duplication], [Workflow]User: "Audit the creative-writing plugin suite"
Action:
1. Load creative-writing plugin context
2. Run all four analysis phases
3. Generate comprehensive report
4. Offer todo conversion
User: "Check usability of testing plugin agents"
Action:
1. Load testing plugin context
2. Run usability analysis only
3. Generate focused report
User: "What agents should we add to the observability plugin?"
Action:
1. Load observability plugin context
2. Run gap analysis only
3. Generate agent recommendations with specs
User: "Find duplicate functionality in the knowledge-base plugin"
Action:
1. Load knowledge-base plugin context
2. Run duplication analysis only
3. Generate consolidation recommendations
This skill complements the grey-haven-plugin-audit skill:
| Skill | Focus | Level |
|---|---|---|
plugin-audit | Structure, frontmatter, deprecations | Technical validation |
suite-audit | Usability, gaps, duplication, workflow | Strategic improvement |
Recommended workflow:
plugin-audit first for technical healthsuite-audit for strategic improvementsSkill Version: 1.0 Last Updated: 2025-01-08
npx claudepluginhub greyhaven-ai/claude-code-config --plugin plugin-auditorAssesses Claude Code plugins by analyzing structure, scoring quality, and generating phased refactoring plans with design maps and task files for parallel agent execution.
Analyzes Claude Code plugins for sub-agent opportunities, identifying skills with verbose output, gaps in agent coverage, and model selection improvements like opus for complex reasoning.
Runs parallel analyzers over plugin manifests, agents, skills, CLAUDE.md, docs, prompts, commands, and hooks; reports HIGH-certainty findings first and optionally applies safe fixes.