Audit AI project configuration and maintainability across five layers.
npx skills add https://github.com/TLOGBen/baransu --skill health-tlogben
Please help me install this Agent Skill. Skill: health Source: https://github.com/TLOGBen/baransu/tree/main/codex/plugins/baransu/skills/health Command: npx skills add https://github.com/TLOGBen/baransu --skill health-tlogben
This Skill provides a comprehensive audit of an AI project's configuration and maintainability, identifying potential issues and providing actionable insights to improve code quality and security.
Run the health skill with the /health command.
High-intent search queries and answers about installing and using this skill.
To audit AI agent configuration for instruction drift and security vulnerabilities, you need a multi-layer audit framework that inspects agent configuration, instruction surfaces, tools, and verifiers to detect risks and provide actionable maintainability insights.
A five-layer audit framework for AI coding maintainability comprehensively evaluates agent configuration, instruction surfaces, tools, verifiers, and overall maintainability to identify potential issues like code rot and improve project reliability.
Yes, you need python, jq, git, and git-lfs installed to run an AI project maintainability audit, as the analysis requires executing various checks and parsing project configuration files to detect risks.
You can check your AI project for code rot and instruction drift by running a comprehensive health audit that analyzes project files and execution surfaces, providing specific recommendations to address identified security and maintainability risks.
The best way to identify security vulnerabilities in AI coding tools is to perform a comprehensive audit of the project's configuration, inspecting tools and verifiers to detect risks and provide actionable steps to improve code quality.
yogogml0035-cpuAutomate harness documentation updates from real-time project analysis.
tw93Audit instruction drift and configuration health in AI engineering projects.
JohnRobert-OlanoBlock destructive git commands in Claude Code before execution.
julianobarbosaReview code for quality, bugs, security issues, and best practices using Codex.
