Transform any codebase into an intelligent knowledge graph. Built-in MCP server with memory and optimized search connects seamlessly with Claude Desktop, Cursor, and other AI assistants. Plus everyday CLI tools for commits, reviews, and releases.
Built-in Model Context Protocol server with persistent memory and optimized search for AI assistants
Smart commit messages, AI code reviews, and automated releases - all from your terminal
Natural language queries with multi-query support across 12+ programming languages
AI-powered relationship discovery with GraphRAG technology
Local-first processing, intelligent vector optimization, handles codebases of any size
Apache 2.0 license, community-driven, transparent development
Octocode provides a production-ready Model Context Protocol server with persistent memory, optimized search, and intelligent caching. Your AI assistants get comprehensive codebase knowledge that improves over time.
AI assistants remember insights, decisions, and patterns across sessions
Multi-query semantic search with intelligent result ranking and caching
Works with Claude Desktop, Cursor, Continue, and any MCP-compatible AI tool
octocode mcp --path /your/project
MCP Server Features
Beyond search and indexing, Octocode provides AI-powered CLI tools for your daily development workflow.
octocode commit --all
AI analyzes your staged changes and generates contextual commit messages following conventional format with proper scope and detailed descriptions.
octocode review --focus security
Get AI-powered code reviews that check for security issues, performance problems, best practices, and maintainability concerns.
octocode release --dry-run
Intelligent release automation with AI-powered version calculation, changelog generation, and semantic versioning based on your changes.
Generate contextual commit messages, get AI-powered code reviews, and automate releases. octocode commit --all analyzes your changes and creates meaningful commit messages following conventional format.
Ask your AI assistant "How does authentication work?" and get comprehensive answers with code examples, relationships, and architecture insights from your MCP-enabled knowledge graph.
Your AI assistant remembers previous conversations, architectural decisions, and code patterns. Build institutional knowledge that persists across team members and projects.
Search with multiple related terms like ["auth", "middleware", "jwt"] for comprehensive results. The MCP server optimizes and caches searches for faster AI assistant responses.
Your data, your control, Apache 2.0 license
Add custom languages, embedding providers, or AI models
Audit the code, run locally, enterprise-ready
Backed by Muvon, enhanced by developers worldwide
curl -fsSL https://raw.githubusercontent.com/Muvon/octocode/main/install.sh | shexport VOYAGE_API_KEY="your-key"octocode index# Start MCP server for AI assistants
octocode mcp --path .
# Try AI-powered git workflow
octocode commit --all
octocode review --focus securitycargo install --git https://github.com/Muvon/octocodeNeed enterprise features, custom integrations, or dedicated support?
Contact Muvon Team