skills.rest
treejh
treejhCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

health

Audit Claude Code six-layer stack configurations and summarize categorized findings with remediation guidance.

Updated Apr 9, 2026
Search Tags:#verification#audit#hooks#configuration#mcp#claude-code#health-check
One-click install
npx skills add https://github.com/treejh/ai-agent-skills --skill health-treejh
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: health
Source: https://github.com/treejh/ai-agent-skills/tree/main/skills/health
Command: npx skills add https://github.com/treejh/ai-agent-skills --skill health-treejh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, python3, and includes scripts (resource) components.

What problem does it solve?

This Skill helps when Claude ignores instructions, behaves inconsistently, hooks malfunction, or MCP servers cause failures by auditing the full six-layer Claude Code configuration stack and surfacing prioritized issues so engineers can fix root causes quickly.

Core Features & Use Cases

  • Six-layer audit: Inspects CLAUDE.md, rules, skills, hooks, subagents, and verifiers to locate misconfigurations and misplaced directives.
  • Automated data collection: Runs a collection script to gather metrics, conversation signals, hooks, MCP settings, and skill frontmatter for analysis.
  • MCP live checks & agent analysis: Verifies MCP server liveness, launches two parallel analyzers (context/security and control/behavior), and classifies findings into Critical, Structural, and Incremental.
  • Use case: Trigger when Claude sessions skip hooks or repeatedly ignore rules, when MCP tools fail, or as part of a periodic configuration health check.

Quick Start

Invoke the health skill by running /health to collect configuration data and generate a prioritized health report.

Frequently Asked Questions about health

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I audit Claude Code configurations for missing hooks or misconfigurations?▼

You can audit Claude Code configurations by running a health check script that inspects the six-layer stack, including CLAUDE.md, rules, skills, hooks, subagents, and MCP servers, to surface prioritized issues and remediation guidance.

Why does Claude ignore instructions and skip configured hooks during sessions?▼

Claude may ignore instructions due to misconfigurations or context pressure within the six-layer stack; an automated health check collects local artifacts and conversation signals to classify these structural issues and identify root causes.

How do I check if MCP servers are causing failures in my Claude Code setup?▼

To check MCP server failures, run a health check that performs live MCP server liveness verification alongside data collection, identifying critical configuration issues that cause tools to malfunction.

What dependencies do I need to run a Claude Code health check?▼

You need jq and python3 installed in your environment to execute the automated data collection script that gathers metrics, hooks, MCP settings, and skill frontmatter for the health audit.

What is the best way to troubleshoot inconsistent Claude Code behavior across projects?▼

The best way to troubleshoot inconsistent behavior is to audit the full configuration stack by launching parallel analyzers for context, security, control, and behavior, which categorize findings into critical, structural, and incremental severities.

Related Skills

Explore More Software Engineering Skills→
andfaniloandfanilo
629

using-streamlit-cli

Manage and run Streamlit apps via command line with configuration and diagnostics.

Community
Intermediate
einargudnigeinargudnig

health

Audit Claude Code six-layer violations and surface misaligned layers.

Community
Advanced
zylos-aizylos-ai
1.1k

activity-monitor

Monitor Claude agent status and restart it after crashes using PM2.

Official
Advanced
47network47network
5

sast-scanner

Scan TypeScript/JavaScript source code for OWASP-aligned security vulnerabilities.

Community
Intermediate