zilliztech/claude-context

Claude Context is a plugin that gives AI coding assistants (like Claude Code or Cursor) the ability to search your entire codebase by meaning rather than by exact text match. When you ask an AI assistant to help with a coding task, it normally only sees the files you have open or the small snippet you paste in. Claude Context changes that by indexing all of your code into a vector database, a kind of database that stores the meaning of text as numbers, then retrieving the most relevant pieces whenever the AI needs them. The connection between the plugin and the AI assistant goes through something called MCP (Model Context Protocol), which is a standard way for tools to plug into AI coding environments. Once configured, the AI can call into Claude Context to find relevant code from across your project without you having to manually point it to the right files. This approach is described as cost-effective because it avoids loading entire folders of code into the AI on every request, which can be expensive when AI services charge per amount of text processed. Instead, only the relevant portions are pulled in. Setting it up requires two external services: a Zilliz Cloud account (which provides the vector database, with a free tier available) and an OpenAI API key (used for the embedding model that converts code into searchable number representations). Configuration involves adding a few lines to a settings file in whichever coding tool you use, and the README covers setup for Claude Code, Cursor, Gemini CLI, Windsurf, and several others. The project is MIT licensed and published by Zilliz, the company behind the Milvus open source vector database. It is aimed at developers working on large codebases who want their AI assistant to have broader awareness of the project without manual file management.