Skip to main content
SHARC (Semantic Hybrid Architecture for Repository Code-search) is a powerful MCP-powered semantic code search tool that helps AI assistants understand and navigate codebases more efficiently.

​
What is SHARC?

SHARC provides semantic code search capabilities to AI assistants like Claude Code, Cursor, and other MCP-compatible tools. Instead of relying on keyword matching or file-by-file exploration, SHARC uses state-of-the-art embeddings to understand the meaning of your code and find relevant results instantly.

​
Key Benefits

10x Fewer Tool Calls

Semantic search finds relevant code directly, eliminating the need for multiple grep/read operations.

33x Less Code to Process

Return only the most relevant code snippets instead of entire files.

15x Faster Results

Get answers in seconds, not minutes. Real-time file watching keeps your index current.

Getting Started

Install SHARC and configure it with your AI assistant in minutes.

MCP Tools

Learn about the 7 MCP tools available for indexing and searching.

Packages

See published NPM packages for the MCP server and code splitter.

Architecture

Understand how SHARC works under the hood.

​
How It Works

​
Get Started

The fastest way to get started is with Claude Code: 
claude mcp add sharc \
  -e SHARC_API_KEY=your_api_key \
  -- npx @sharc-code/mcp@latest
Then index your codebase and start searching: 
> Index this codebase for semantic search

● index_codebase (MCP)
  ⎿ Indexed 2,450 chunks from 342 files in 18.3s

> How does authentication work in this project?

● search_code (MCP)
  ⎿ Found 3 results for query: "authentication"