# AI Guard > AI Guard (eslint-plugin-ai-guard) is an open-source ESLint plugin, CLI and set of agent hooks that catches risky patterns AI coding tools often produce in JavaScript and TypeScript and checks npm packages before they are installed. Version 1.4.7: 27 lint rules in 4 presets, hooks for Claude Code, Codex and Cursor, and an MCP server with 9 tools. AI Guard does not detect AI-written code. It reports risky patterns that AI tools often produce, wherever they came from. Install: `npm install --save-dev eslint-plugin-ai-guard`, then `npx ai-guard run`. Every fact on this site is also in [facts.json](https://getaiguard.dev/facts.json). Links below go to markdown versions of each page; drop `.md` for the HTML page. ## Getting started - [Getting started](https://getaiguard.dev/docs/getting-started.md): install, first scan, presets, agent hooks, package checks, CI - [What is AI Guard?](https://getaiguard.dev/about.md): what it does, who it is for, what it is not ## Package checks - [Package checks](https://getaiguard.dev/package-checks.md): blocked versus warned, the fresh-release check, privacy, commands - [What is slopsquatting?](https://getaiguard.dev/guides/slopsquatting.md): AI-invented package names and how the check refuses them - [How to check an npm package before installing it](https://getaiguard.dev/guides/check-npm-package.md): npm view, then npx ai-guard verify-deps, with real output - [Release cooldowns](https://getaiguard.dev/guides/release-cooldowns.md): min-release-age and its equivalents in pnpm, Yarn and Bun ## Agent hooks - [Guardrails for Claude Code, Codex and Cursor](https://getaiguard.dev/guides/ai-agent-guardrails.md): what each agent gets, the finish gate, the --no-verify guard, tests - [Claude Code](https://getaiguard.dev/docs/claude-code.md): project hooks and the Claude Code plugin - [Integrations](https://getaiguard.dev/integrations.md): Codex, Cursor, pre-commit gate, Oxlint, MCP server, GitHub Action - [Checks on the agent's change](https://getaiguard.dev/docs/verify.md): verify, integrity, test integrity, scan-agent-config ## Rules - [Rule catalog](https://getaiguard.dev/rules): all 27 rules with their level in each preset - [Why no automatic fix for empty catch blocks and floating promises](https://getaiguard.dev/guides/why-no-autofix.md): a fix that hides the finding defeats the purpose - [AI Guard and ESLint](https://getaiguard.dev/compare/eslint.md): a plugin that extends ESLint, not a replacement ## Results - [Results on real projects](https://getaiguard.dev/results.md): measurements, method, limits; source docs/evidence.md ## MCP server - [MCP server](https://getaiguard.dev/docs/mcp-server.md): 9 tools over local STDIO; MCP Registry id io.github.ai-guard-dev/ai-guard ## GitHub Action - [GitHub Actions and SARIF](https://getaiguard.dev/docs/github-action.md): workflow, inputs, outputs, Code Scanning upload ## Changelog - [CHANGELOG.md](https://github.com/ai-guard-dev/eslint-plugin-ai-guard/blob/main/CHANGELOG.md): every release ## Source - [GitHub repository](https://github.com/ai-guard-dev/eslint-plugin-ai-guard): source, docs/, issues (MIT) - [npm package](https://www.npmjs.com/package/eslint-plugin-ai-guard): eslint-plugin-ai-guard ## Optional - [llms-full.txt](https://getaiguard.dev/llms-full.txt): the main documentation in one file - [AI Guard and npm audit](https://getaiguard.dev/compare/npm-audit.md): what each checks, and why to use both - [CLI reference](https://getaiguard.dev/docs/cli.md): every command and exit code