st3v3nmw/sourcerer-mcp

Overview

A Model Context Protocol server that enables semantic code search and navigation, reducing token waste by indexing codebases and retrieving only relevant snippets for LLM tools.

Best for

Best for
Developers building MCP-enabled code assistants who need efficient semantic code retrieval

Use cases

• Performing natural language queries against codebases for LLM-powered code analysis
• Navigating large repositories without sending full context to AI assistants
• Integrating semantic search into MCP-compatible development environments

How to use

Tools exposed

• sourcerer

Tested with

Claude Code

Example client config

{\n  "mcpServers": {\n    "sourcerer": {\n      "command": "sourcerer",\n      "env": {\n        "OPENAI_API_KEY": "your-openai-api-key",\n        "SOURCERER_WORKSPACE_ROOT": "/path/to/your/project"\n      }\n    }\n  }\n}

Notes

A Model Context Protocol server that enables semantic code search and navigation, reducing token waste by indexing codebases and retrieving only relevant snippets for LLM tools.

114 stars on GitHub. Last updated 2025-11-10. Licensed MIT.

Use cases

• Performing natural language queries against codebases for LLM-powered code analysis
• Navigating large repositories without sending full context to AI assistants
• Integrating semantic search into MCP-compatible development environments

Pros

• Reduces token consumption by delivering only contextually relevant code to LLMs
• Semantic search improves accuracy over simple keyword matching
• Single Go binary is fast and easy to deploy

Cons

• Relatively new with limited community adoption and fewer than 150 stars
• Requires an MCP client or framework to use, limiting standalone utility
• May need to be self-hosted or configured per project

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.