# Use Confluent Documentation with AI Tools Confluent documentation integrates with AI-powered workflows and IDEs. This enables agents to maintain persistent access to accurate, current documentation throughout their interaction. This page describes the formats and tools Confluent provides for AI workflows, from copying Markdown to running a documentation Model Context Protocol (MCP) server. ## Markdown source content Confluent documentation pages are available in both HTML and Markdown formats. Language models and AI-powered tools parse Markdown more efficiently than HTML. Use the **Copy Markdown** or **Open as Markdown** buttons to share a documentation page’s content with AI agents or provide it as context in development tools and IDEs for faster, more accurate responses. ## MCP server for documentation You can use the [Confluent MCP server](https://github.com/confluentinc/mcp-confluent) to search and retrieve live Confluent documentation directly through an MCP server. The server runs locally on your machine and works with any MCP-compatible client, such as Claude Code, Cursor, and VS Code. The MCP server ensures that your AI-powered tools read the latest version of the Confluent documentation. The server includes two documentation tools: - `search-product-docs` searches across `docs.confluent.io`, `developer.confluent.io`, and `support.confluent.io`. - `get-product-doc-page` retrieves the full Markdown content of a documentation page. To use the documentation tools, generate a starter configuration file and start the server: ```bash npx @confluentinc/mcp-confluent --init-config npx @confluentinc/mcp-confluent --config ./config.yaml ``` For full setup, authentication, and configuration details, see [Build with the Open-Source MCP Server for Confluent](https://docs.confluent.io/cloud/current/ai/ai-tools/open-source-mcp-server.html). ## Compressed documentation index for coding agents Confluent provides a compressed documentation index that you can add to your project rules file, for example `CLAUDE.md` or `AGENTS.md`. This index provides a persistent context to AI coding agents by mapping documentation topics to their source files. This can reduce hallucinations and ensure agents reference the live documentation rather than relying on training knowledge. Place the index file in your project root. [`Download the Confluent documentation index`](../.hidden/docs-common/home/includes/docs-llm-index.txt) ## llms.txt and llms-full.txt standard files [llms.txt](https://docs.confluent.io/llms.txt) and [llms-full.txt](https://docs.confluent.io/llms-full.txt) use a standard format that makes documentation discoverable to AI tools and language models (LLMs). These files list key documentation pages and resources in a format that LLMs can easily parse and reference. ## Context7 integration Unlike the Confluent MCP server, Context7 is a third-party MCP server that aggregates documentation from many sources. [Context7](https://context7.com/) is an MCP server that provides up-to-date documentation directly into your AI-powered Integrated Development Environment (IDE), such as VS Code and Cursor. Confluent registers its documentation with Context7 and updates it regularly.