# ondocs > Make your docs answerable by people and by agents, from one static index: a cited Ask box, an MCP server and llms.txt, on your own function and key, with no vector database and no vendor. Agent traffic costs you no model tokens. Plugins for Docusaurus, Astro and Starlight, React components for Next.js, and a script tag for any other site. Install with `npm i ondocs`. Source: https://github.com/dgesteves/ondocs. Each page links to its Markdown copy: its URL plus `.md`. Every page in one file: https://ondocs.vercel.app/llms-full.txt Agents can search these docs over MCP, with search, fetch and list_pages tools: https://ondocs.vercel.app/api/mcp ## Get started - [Introduction](https://ondocs.vercel.app/docs/introduction.md): What ondocs is, who it is for, and when it is the wrong tool. - [Getting started](https://ondocs.vercel.app/docs/getting-started.md): Install the package, build the index, deploy the endpoint and add the dialog. - [Try it locally](https://ondocs.vercel.app/docs/local-development.md): Run the whole pipeline on your machine with mock mode and ondocs dev, no API key. ## Integrations - [Docusaurus](https://ondocs.vercel.app/docs/docusaurus.md): The Docusaurus 3 plugin, which indexes the built site and adds the dialog. - [Astro and Starlight](https://ondocs.vercel.app/docs/astro.md): The Starlight plugin and the Astro integration, which index the built site and add the dialog. - [VitePress](https://ondocs.vercel.app/docs/vitepress.md): The VitePress plugin, which indexes the built site and adds the dialog, with one change to the config. - [Next.js and React](https://ondocs.vercel.app/docs/nextjs.md): Index with the CLI, mount createAskHandler as a route handler, and render AskDialog. - [Script tag](https://ondocs.vercel.app/docs/script-tag.md): Add the dialog to Hugo, Jekyll, Eleventy, MkDocs or plain HTML with one script tag. - [MCP server](https://ondocs.vercel.app/docs/mcp.md): Serve the same index to coding agents and AI apps as an MCP server, with search and fetch tools, and no model call on your side. - [llms.txt and Markdown pages](https://ondocs.vercel.app/docs/llms-txt.md): Write llms.txt, llms-full.txt and a Markdown copy of every page at build time, from the pages the index is built from. - [Writing an integration](https://ondocs.vercel.app/docs/writing-an-integration.md): Build a plugin for another docs framework from ondocs/integration, the helpers the Docusaurus, Astro, Starlight and VitePress plugins are made of. ## Guides - [Indexing content](https://ondocs.vercel.app/docs/indexing.md): What gets indexed, how files become URLs, chunking, excluding pages and the --check mode. - [Model providers](https://ondocs.vercel.app/docs/model-providers.md): Bring your own key for OpenAI, Anthropic, Google, xAI, AI Gateway or any AI SDK provider, and configure the embedding model. - [Deploying](https://ondocs.vercel.app/docs/deployment.md): Write the endpoint for Vercel, Netlify, Cloudflare or GitHub Pages with one command, where the index lives, cold starts and cost. - [Workers AI, without an API key](https://ondocs.vercel.app/docs/workers-ai.md): A Cloudflare Worker that answers for any static docs site with Workers AI, the models and their cost, and how large a site fits. - [Rate limits and security](https://ondocs.vercel.app/docs/security.md): Rate limiting, client IPs, CORS, prompt injection and what data leaves your servers. - [Bot check with Turnstile](https://ondocs.vercel.app/docs/bot-challenge.md): Ask visitors' browsers to pass a Cloudflare Turnstile check before each question, so a fleet of bots cannot spend your model budget. - [How retrieval works](https://ondocs.vercel.app/docs/retrieval.md): BM25, cosine similarity, reciprocal rank fusion and the relevance gate that says "I don't know". - [Measure and improve answers](https://ondocs.vercel.app/docs/quality.md): Score retrieval in CI with ondocs eval, collect readers' feedback, and keep a log of the questions your docs do not answer. ## Reference - [Package API](https://ondocs.vercel.app/docs/api.md): The package's imports, the core functions for building and searching an index, and what each entry point exports. - [CLI and config file](https://ondocs.vercel.app/docs/cli.md): Every flag of ondocs index, init, dev and eval, and the config module. - [The ask endpoint](https://ondocs.vercel.app/docs/ask-endpoint.md): createAskHandler options, the request and response, errors and the streaming protocol. - [The ask dialog](https://ondocs.vercel.app/docs/ask-dialog.md): AskDialog props, the launcher button, shortcuts, theming, accessibility and the useAsk hook. - [The index file](https://ondocs.vercel.app/docs/index-format.md): What ask-index.json holds, why vectors are int8, and how the file stays small and diffable. - [Benchmarks](https://ondocs.vercel.app/docs/benchmarks.md): Index size, cold start, memory and query latency at 1,000, 10,000 and 50,000 chunks. - [Limits and trade-offs](https://ondocs.vercel.app/docs/limits.md): Where ondocs stops being the right tool, and what it gives up for simplicity. - [FAQ](https://ondocs.vercel.app/docs/faq.md): Short answers to the questions people ask first, with links to the details. - [Migrating from ask-my-site](https://ondocs.vercel.app/docs/migrating.md): ask-my-site is now ondocs. What was renamed, what stays, and the commands to move a site over.