A linter and LSP that validates configuration files for AI coding assistants like Claude Code, Cursor, and Copilot.
agnix is a specialized linter and language server that validates configuration files for AI coding assistants. It catches syntax errors and misconfigurations in files like CLAUDE.md, SKILL.md, and MCP configs before they cause AI tools to silently fail or produce incorrect outputs. The project aims to eliminate 'almost right' configurations by providing rigorous, automated validation across multiple tools.
Developers and teams using AI coding assistants like Claude Code, Cursor, GitHub Copilot, Kiro, and others who need to ensure their agent configuration files are syntactically correct and adhere to tool-specific specifications.
Developers choose agnix because it offers comprehensive, multi-tool validation with 385 rules sourced from official specifications and real-world breakage patterns, preventing silent failures. It provides automated fixes with confidence levels and integrates directly into editors and CI/CD pipelines, unlike generic linters that don't understand AI agent configuration formats.
The missing linter and lsp for AI coding assistants. Validate CLAUDE.md, AGENTS.md, SKILL.md, hooks, MCP. Plugin for all major IDEs included, with autofixes.
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
With 385 rules covering Claude Code, Cursor, Copilot, Kiro, and more, agnix catches syntax errors and misconfigurations across the entire AI coding stack, preventing silent failures as highlighted in the README's tool support table.
Plugins for VS Code, JetBrains, Neovim, and Zed provide real-time linting and autofix capabilities directly in the editor, as demonstrated in the installation and usage sections.
The CLI and editors offer fixes categorized by confidence (safe, unsafe), allowing developers to apply corrections with awareness of risks, shown in the --fix-safe and --fix-unsafe options.
Includes a pre-built GitHub Action for integrating validation into continuous integration pipelines, ensuring agent configs are checked before deployment, as specified in the README.
The README notes 'new rules and tool support ship constantly,' which can lead to frequent updates, potential breaking changes, and the need for users to stay updated to avoid incompatibilities.
Adding agnix introduces another linter to the workflow; for teams using only basic, well-documented AI configs, this might be unnecessary complexity that slows down development.
agnix only validates configs for the listed tools; custom or emerging AI assistant formats are not covered until explicitly added, as seen in the supported tools table, which may lag behind rapid tool evolution.