What the tech

Search

Browse topics

/ opens search from anywhere. Esc closes it.

Tools
What the tech
Get the newsletter

health: a Claude Code skill by Tw93

For Claude Code, Codex or Pi users whose agent ignores instructions or whose hooks and MCP servers misbehave. Skip it if you are chasing an app bug or want a pull request review.

The skill's text changed on Nov 10, 2026, after we last read it. Our description may be behind.

Copy install command View the source
Score
70of 100
Rank
#177of 239
Installs
19.1Kon skills.sh
Security
Clear 4 scanners
Updated
2026
Context
~90tokens always

Install

npx skills add https://github.com/tw93/waza --skill health

Run it in your terminal. View the source

Using the Claude app rather than Claude Code? You can add a skill there too. How to add it in the app

What does health do?

health is a skill that audits a project's AI agent setup and writes a report-only health check. Claude first runs the bundled collect-data.sh script, or a PowerShell launcher on Windows, which gathers instruction files, settings, hooks, MCP servers, skills and up to three recent Claude and Codex sessions for the project. It then checks deny-lists for credential folders, permission-skip flags, unpinned third-party skills, memory hygiene and drift between AGENTS.md and CLAUDE.md. Findings land under Critical, Structural or Incremental, each with the file, a reason and a copy-pasteable fix. A deep mode streams every past session and runs inspector sub-agents, and Claude warns before using it.

When should you use health?

Use it when

  • Claude keeps ignoring rules in CLAUDE.md or AGENTS.md
  • A hook or MCP server stopped working
  • You want permissions and skills checked for risky settings

Skip it when

  • The problem is a bug in your app code
  • You want a pull request reviewed
  • You do not want past sessions read

What it needs and what it can touch

You need

  • Bash on macOS or Linux, or PowerShell on Windows
  • Python 3 for the full set of checks

What it can do

  • Runs its bundled shell and Python scripts
  • Reads agent settings, instruction files and local session histories
  • Starts sub-agents in deep mode

From reading its instructions, not from running it.

Why it scores 70

Strongest on upkeep (100 of 100), weakest on adoption (43).

  • Adoption35 points 43
  • Upkeep20 points 100
  • Trust20 points 65
  • Quality15 points 100
  • Reader answers10 points Not enough answers yet, so this does not count

How the formula works

Is health safe to install?

Every scanner that checked it passed it. Scanners often miss what other scanners catch, and a skill can change after a scan. Read its source before you install it.

ScannerResultDate
Gen Agent Trust HubPassedNov 10, 2026
SocketPassedOct 29, 2026
SnykLow riskNov 10, 2026
ZeroLeaksPassedApr 16, 2026

Details

Installs
19,106
Repository stars
7,548
Last repository change
Tokens always loaded
about 90
Tokens when used
about 5,115
License
MIT
Data confidence
High

Alternatives to health

All skills for this job