grepai

AI-Powered Semantic Code Search & Repository Intelligence

A privacy-first, open-source developer tool that enables developers and AI coding agents to understand, navigate, and analyze entire codebases using natural language.

The application combines semantic vector search, keyword-based retrieval, AST analysis, and dependency graphs to locate relevant code, trace function calls, identify references, and understand architectural relationships.

Core capabilities:

• Natural-language and hybrid semantic code search
• Automatic repository indexing with real-time file synchronization
• Function call graphs, dependency tracing, and impact analysis
• Multi-language, multi-repository, and Git worktree support
• Local-first execution with optional cloud embedding providers
• MCP integration with Claude Code, Cursor, Codex, and other AI agents
• CLI, API, and optional visual dashboard
• Fast, incremental indexing with configurable vector storage

Goal: Give developers and AI agents accurate, context-aware access to codebase knowledge, reducing exploration time, unnecessary token consumption, and refactoring risks while keeping source code private.