The first audit tool for Claude Code. Maps every plugin, skill, command, hook, and MCP server — scores 8 dimensions — generates a prioritized optimization plan with copy-paste config fixes.
Built by the team that optimized their own setup from 24 plugins down to 13.
Every Claude Code power user hits the same wall. You install plugins “just in case.” MCP servers connect but never get used. Custom commands duplicate plugin skills. MEMORY.md silently overflows its 200-line limit.
Sessions feel slow and unfocused
Context window bloated with 100+ phantom MCP tools
"Which code review tool do I use?"
5-7 overlapping capabilities with no decision tree
Claude forgets context between sessions
MEMORY.md exceeded 200-line limit — silently truncated
Hooks conflict with workflow rules
Auto-commit to main while rules say "always branch"
You don't have a capability problem. You have an organization problem.
/auditReads all config files — settings, plugins, commands, hooks, MCP servers, memory.
Counts every capability. Built-in tools. Plugin skills. Custom commands. Subagents. MCP tools.
Rates your setup across 8 dimensions on a 0-10 scale with justification.
Maps every overlap — code review, web search, deployment — flags areas needing decision trees.
Prioritized recommendations (P0 through P3) with copy-paste config diffs.
Full markdown report saved to your project. Composite score with letter grade.
One confirmation, instant optimization. Quick wins applied in seconds.
Each dimension rated 0-10 with explanation of what would make it a 10.
Coverage
Right plugins for your stack
Redundancy
How many tools do the same thing
Config Hygiene
Conflicts, stale config, missing CLAUDE.md
Memory Health
MEMORY.md size and organization
Workflow
Full lifecycle coverage
Command Quality
Documentation, overlap, business value
MCP Utilization
Connected servers actually used
Security
Permissions, secrets, safety hooks
Grading: A (90-100%) Optimized · B (75-89%) Good · C (60-74%) Functional · D (40-59%) Underutilized · F (0-39%) Needs setup
We ran the first audit on our own setup. Here's what changed.
| Metric | Before | After |
|---|---|---|
| Plugins enabled | 24 | 13 |
| MEMORY.md lines | 247 (truncated) | 78 |
| Code review methods | 7 | 3 |
| Phantom MCP tools | 39 | 0 |
| Auto-commit conflicts | Yes | Fixed |
| Composite score | C (62%) | B+ (84%) |
The plugin is free. Pay only if you want hands-on optimization.
$0
Install the plugin, run the audit, get the full report.
$79
We apply the optimizations for you. Before/after comparison included.
$199
Full optimization + custom skills and hooks + 30-minute walkthrough call.
/audit/audit json to generate exportThe audit is read-only by default. It only applies changes when you explicitly approve each one.
Yes. It detects your stack (Node, Python, Rust, Go, Ruby) and scores coverage accordingly. The scoring adapts to what matters for your project.
Even a minimal setup benefits. The audit checks workflow gaps, memory health, and security — not just plugin count. Some of the biggest wins come from missing tools, not removing extras.
Yes. Run /audit monthly to track your score over time. The JSON export enables before/after comparison across audits.
No. Everything runs locally in your Claude Code session. For Standard and Premium tiers, you choose what to share with us.
290 capabilities. You're using a fraction of them. The rest are costing you context window, cognitive load, and missed potential.