memov

MemoV (VibeGit) acts as a specialized memory layer designed for AI coding agents and developers who want to maintain a granular, immutable history of their development sessions. Unlike standard Git, which records snapshots of file states, MemoV preserves the intent behind every change, mapping user prompts, AI responses, and corresponding code diffs into a structured timeline. This allows developers to audit why a specific code change was made, jump between historical snapshots, and experiment with different AI-driven branches without cluttering the primary version control repository.

It is an essential tool for those building with agents like Claude Code or Cursor, providing a bridge between raw file modifications and the conversational context of an AI interaction. By keeping a local .mem shadow repository, it ensures your main .git workflow remains clean while adding a sophisticated layer of 'vibe' engineering that tracks the evolution of your codebase through every request.

• Automatically captures prompts, AI responses, and resulting code changes via a unified CLI or MCP interface.

• Supports time-travel debugging and branch exploration, allowing developers to revert or jump to any previous interaction state.

• Keeps the core .git repository free of transient development snapshots while maintaining full context via a local .mem directory.

• Provides a visual interface (Web UI) for browsing the history of coding sessions locally at localhost:38888.

• Offers RAG-ready features including semantic search and codebase validation when optional dependencies are installed.

• Seamless integration with modern AI coding environments like Claude Code, Codex, and VS Code.

• Ensure you have the memov package installed via pip or official scripts to access the mem CLI.

• Use the init command in your project root to initialize the shadow tracking layer.

• Always record sessions using the snap command or through the configured MCP server tools for the best integration.

• The tool is private-first, running entirely locally without a central database, ensuring sensitive codebase information remains secure.

• Utilize .memignore to exclude non-essential files and keep the repository history focused on relevant changes.

• For developers requiring advanced searching or debugging features, use pip install memov[rag] to enable vector-based lookup capabilities.