config-lint

What problem does it solve?

Manually validating and fixing fragmented AI assistant configuration files (Claude Code skills, hooks, MCP servers, GitHub Copilot instructions, and CLAUDE.md files) for compliance with official Anthropic and GitHub best practices is time-consuming and error-prone, especially for team-shared projects where configuration consistency is critical.

Core Features & Use Cases

• Unified multi-linter workflow: Runs 5 specialized linters in a single pass to cover all common AI assistant config types, eliminating the need to use separate tools for skills, hooks, MCP, Copilot instructions, and CLAUDE.md files.
• Auto-remediation support: Automatically fixes common structural issues (such as deprecated hook output formats or missing required frontmatter fields) when run with the --fix flag.
• Use Case: A developer preparing a new Claude Code project for team collaboration can use this skill to audit all configuration files before committing them to the repository, ensuring they meet governance standards and avoid common pitfalls like hardcoded secrets or invalid hook matchers.

Quick Start

Use the config-lint skill to validate and fix all AI assistant configuration files in your current project for compliance with official best practices.