Indexes and serves documentation from various sources via the Model Context Protocol and REST API to provide context for AI assistants.
Sponsored
Context is a self-hosted server designed to transform your documentation into a searchable knowledge base accessible by AI assistants. It works by parsing content from diverse sources like MDX, Markdown, and OpenAPI specifications, then semantically chunking and embedding it using OpenAI. This processed information is stored in a Pinecone vector database, enabling AI assistants to query and retrieve relevant context efficiently through the Model Context Protocol (MCP) and a REST API.
Key Features
012 GitHub stars
02Serves content via MCP and REST API
03Stores vectorized data in Pinecone vector database
04Utilizes OpenAI for content embeddings
05Indexes documentation from multiple sources (MDX, Markdown, OpenAPI)
06Self-hosted Model Context Protocol (MCP) server
Use Cases
01Building searchable knowledge bases for AI assistants
02Providing AI with context from documentation platforms like Mintlify, Fumadocs, and Docusaurus
03Enabling AI assistants to query API reference specifications (OpenAPI/Swagger)