Show HN: Agnix – lint your AI agent configs (Claude.md, skills, MCP, hooks) (github.com)

Agnix has been launched as a linter specifically designed for AI agent configurations, targeting essential components such as skills, hooks, memory, and plugins within coding stacks. It addresses a critical pain point for developers: misconfigurations that often lead to minimal functionality, with research indicating that skills can fail to trigger if even slightly misconfigured. The tool aims to eliminate this frustration—cited by 66% of developers—as "almost right" outcomes amplify incorrect configurations that AIs may learn from. The Agnix linter incorporates 156 validation rules derived from official specifications and extensive real-world testing. It offers functionalities like automatic fixing of configuration errors across various tools, including Claude Code, GitHub Copilot, and Cursor, allowing developers to streamline their workflows. With support for multiple platforms and IDEs, this tool simplifies the validation process by ensuring that configurations adhere to the required formats, preventing silent failures when utilizing multi-tool stacks. Agnix is positioned as a crucial asset for enhancing the reliability and effectiveness of AI agents in development.