claudit
Read-only health audit of your machine and Claude Code setup, ranked by impact and compared with your last audit.
$7
Works with the AI Agents you already use
claudit
Example session with this skill installed
claudit
- Read your context and instructions
- Compiled the claudit
Headline: 2 risks, 5 friction items, 4 missing tools. Since last audit: 3 fixed, 4 still open, 2 new. Top actions: 1) an MCP server has a literal API key in ~/.claude.json headers (value redacted), 2) pyright is 14 months old, 3) a skill references a script that no longer exists. Report: ~/claudit-reports/2026-11-16 claudit results.md
Connects securely to your tools. The creator never sees your data.
About this skill
Ask "is my setup in good shape?" and get a ranked, dated report instead of a guess.
What it checks
- Command-line tools: core tools (
git,gh,rg,fd,jq,ast-grep,semgrepand more), language runtimes and which one wins onPATH, language servers, cloud CLI logins, package manager health,PATHproblems and secrets exported in plain text. - Claude Code configuration: MCP servers (missing commands, literal credentials, failed connections), hooks and their per-call overhead, permission rules that are too broad or still prompt, plugins and marketplaces, skills that reference missing paths,
CLAUDE.mdrules that name tools you don't have, stale auto-memory and disk use.
How it works
- Two read-only subagents scan in parallel on a cheaper model.
- Findings are ranked: blocks Claude or carries risk, costs time on every task, missing tools, cleanup.
- Each item is compared with your previous report as Fixed, Still open, New or Accepted as is, so you don't get nagged about things you decided to keep.
- A dated Markdown report with checkboxes is saved, and chat gets only the headline, counts, top three actions and the path.
It changes nothing except writing the report. Secrets are always redacted. When you want to fix something, it takes one item at a time and waits for your approval.
Audits only; never installs, edits config or runs auth flows on its own. Fixes happen one item at a time with your approval. Assumes Claude Code's file layout.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed today
Requires
Permissions
File Scopes
Read-only apart from the new report. Secret values are never printed.