Introduction

ck (seek) finds code by meaning, not just keywords. It’s grep that understands what you’re looking for — search for “error handling” and find try/catch blocks, error returns, and exception handling code even when those exact words aren’t present.

What is ck?

ck is a hybrid code search tool that combines the familiarity of grep with the intelligence of modern AI embeddings. It’s designed for:

• AI agents that need reliable code search capabilities via MCP protocol
• Developers who want to find code by what it does, not what it’s named
• Teams exploring large codebases and understanding unfamiliar code
• Code reviewers finding related code and patterns across files

Key Capabilities

🔍 Semantic Search

Find code by concept, not keywords:

ck --sem "error handling" src/
ck --sem "authentication logic" src/
ck --sem "database connection pooling" src/

⚡ grep Compatibility

All your muscle memory works:

ck -n "TODO" *.rs
ck -R -i "fixme" .
ck -l "error" src/

🎯 Hybrid Search

Best of both worlds:

ck --hybrid "connection timeout" src/

🤖 AI Integration

Built-in MCP server for Claude Desktop, Cursor, and other AI tools:

ck --serve

How It Works

1. Indexing – ck automatically creates and maintains semantic indexes of your code
2. Embedding – Uses local AI models (BGE, Nomic, Jina) to understand code semantics
3. Search – Finds semantically similar code chunks using vector similarity
4. Results – Returns familiar grep-style output with optional relevance scores

Why ck?

vs. grep/ripgrep

• ✅ Understands code meaning, not just text patterns
• ✅ Finds related code even with different terminology
• ✅ Maintains full grep compatibility for keyword search
• ✅ Automatic smart file filtering (.ckignore)

vs. IDE search

• ✅ Works across entire codebase, not just open files
• ✅ Command-line friendly for scripts and automation
• ✅ Semantic understanding beyond symbol search
• ✅ AI agent integration via MCP

vs. AI code search services

• ✅ 100% offline — no code leaves your machine
• ✅ No API keys or subscriptions required
• ✅ Fast local inference
• ✅ Privacy-first design

Design Philosophy

• Drop-in compatibility – Works like grep, enhances where needed
• Automatic everything – Index management, updates, model downloads
• Privacy-first – Everything runs locally, no telemetry
• Performance matters – Fast indexing, sub-second queries
• AI-native – Built for both humans and AI agents

Next Steps

• Get started with installation
• Learn basic usage patterns
• Explore MCP integration for AI agents
• Read the CLI reference for all options