aws-documentation
Overview
The aws-documentation MCP server is a Model Context Protocol (MCP) server that provides AI assistants and agents with direct access to AWS’s official documentation content through a structured MCP interface. It enables workflows that search, fetch, and recommend AWS documentation pages — returning markdown-formatted content and relevant recommendations — so models can cite accurate AWS technical details and best practices on-demand without manual lookup.
Transport
stdio
Tools
Key Capabilities
- Fetch AWS documentation pages and convert them into structured markdown for easy AI ingestion and citation.
- Perform context-aware documentation searches across the AWS Docs corpus using keywords and filters.
- Provide content recommendations — related or complementary documentation — to help agents explore topics deeply.
- Support regional documentation partitions (e.g., AWS China) by exposing lists of available services in those partitions.
How It Works
The aws-documentation MCP server runs as an MCP service that listens for requests from AI clients configured with MCP support (e.g., Claude Desktop, Cursor, VS Code). When a tool like read_documentation is invoked with a specific AWS Docs URL, the server fetches the HTML page from docs.aws.amazon.com, converts it to markdown, and returns it as structured text the model can reason about.
For search_documentation, the server uses AWS’s official documentation search API to look up matching pages based on a search phrase and optional filters, returning structured results for the assistant to choose from. The recommend tool analyzes a given documentation page to surface related pages that are often relevant to the same topic.
This design makes AWS Docs accessible as first-class context for AI workflows — reducing hallucinations and increasing accuracy by grounding answers in authoritative AWS content.