Overview
Relevant source files
This document provides a high-level introduction to Claude Context: its purpose, architecture, and primary use cases. For detailed information about specific subsystems, see the child pages listed in the table of contents.
For getting started with configuration and setup, see Quick Start Guide.
For architectural details and component interactions, see System Architecture.
For core concepts and terminology definitions, see Key Concepts and Terminology.
Purpose and Scope
Claude Context is a semantic code search system designed to provide AI coding assistants with deep contextual understanding of large codebases. It enables AI agents to discover relevant code across millions of lines without requiring multi-round discovery or loading entire directories into context.
This system addresses two fundamental problems:
- Context Limitation: AI assistants have limited context windows and need efficient methods to find relevant code README.md20-21
- Cost Efficiency: Loading entire codebases into AI context for every request is prohibitively expensive; Claude Context efficiently stores your codebase in a vector database and retrieves only related snippets README.md22-23
Sources: README.md18-23
What is Claude Context
Claude Context is a monorepo-based code indexing and search system that integrates with AI coding environments through the Model Context Protocol (MCP) README.md18-20 It combines three core technologies:
The system provides a core library (@zilliz/claude-context-core) with three primary integration packages defined in the monorepo:
- MCP Server (
@zilliz/claude-context-mcp) - Primary integration for AI coding assistants like Claude Code package.json10 packages/mcp/README.md1-4 - VS Code Extension (
semanticcodesearch) - IDE-integrated semantic search package.json9 - Chrome Extension - Browser-based code indexing for GitHub package.json4
Sources: README.md18-23 package.json1-24 packages/mcp/README.md1-15
Component Architecture
The following diagram shows the relationship between core components and their actual class implementations across the monorepo:
Component to Code Entity Mapping
Sources: package.json8-10 packages/mcp/README.md37-150 README.md18-31
Core Workflow: Indexing and Search
The following diagram traces the data flow through actual functions and classes during the standard MCP lifecycle:
Indexing and Search Code Flow
Sources: README.md18-31 packages/mcp/README.md186-210 package.json8-10
Primary Use Cases
1. MCP Integration with AI Coding Assistants
Claude Context provides a standardized Model Context Protocol interface for AI agents to interact with your codebase.
Supported AI Environments:
- CLI Agents: Claude Code, OpenAI Codex CLI, Gemini CLI, Qwen Code README.md69-150
- IDE Extensions: Cursor, Windsurf, VS Code (via Cline/Roo Code), Void README.md153-220
- Other Platforms: Cherry Studio, Augment, Zencoder README.md210-220
2. Direct API Usage
Developers can use the Context class from @zilliz/claude-context-core directly for programmatic indexing and search in custom Node.js applications package.json8
Deployment Topologies
Claude Context supports various deployment models to balance performance with data privacy:
Sources: README.md37-54 packages/mcp/README.md30-150
Next Steps
For detailed information about specific subsystems:
- Architecture Details: See System Architecture for monorepo structure and component interactions.
- Getting Started: See Quick Start Guide for step-by-step setup instructions.
- Core Concepts: See Key Concepts and Terminology for definitions of semantic search, embeddings, and hybrid search.
Sources: README.md1-15
Refresh this wiki
This wiki was recently refreshed. Please wait 6 days to refresh again.