claude-context-enhanced

Claude Context Enhanced is a stability-focused MCP plugin that adds semantic code search capabilities to Claude Code and compatible agents. It provides a hybrid search experience combining BM25 and dense vector search, incremental indexing, and automatic Milvus flush to ensure up-to-date results. The available tools allow you to index a codebase, search indexed code, clear an index, and check indexing status, enabling fast, scalable semantic search across large repositories. To use it, configure your MCP environment with the required API keys, then add the MCP server using the CLI provided in the repository documentation. Once added, you can index your codebase, run semantic searches with natural language prompts, and manage indexing status from your development workflow.

Prerequisites:

• Node.js 20.x (>= 20.0.0 and < 24.0.0) installed on your machine
• npm or pnpm for package management

Installation steps:

1. Install Node.js if not already installed (select an LTS build compatible with this MCP, Node 20.x).
2. Open a terminal and ensure Node and npm are available:
   • node -v
   • npm -v
3. Install the MCP package (via npx as described in the configuration) or install locally:
   • npm install -g claude-context-enhanced@latest # if you prefer a global install (if supported by the package)
4. Prepare your environment variables (OpenAI API key and Milvus token) as described in the configuration:
   • OPENAI_API_KEY=sk-...
   • MILVUS_TOKEN=your-zilliz-cloud-api-key
5. Add the MCP server using the CLI (as shown in the README), for example: claude mcp add claude-context-enhanced
   -e OPENAI_API_KEY=sk-your-openai-api-key
   -e MILVUS_TOKEN=your-zilliz-cloud-api-key
   -- npx claude-context-enhanced@latest

Tips and notes:

• Ensure Node.js version compatibility (20.x) to avoid runtime issues with claude-context-enhanced.
• Keep your Milvus token secure; do not expose it in source control.
• The MCP server relies on a vector database; configure Zilliz Cloud or your chosen backend as outlined in the documentation.
• If you encounter issues with snapshots or indexing, verify that your environment meets the prerequisites and that the API keys are set correctly.
• The integration supports automated Milvus flush after indexing to improve data availability; monitor health checks via the MCP management tooling.