The problem: tools like Claude Code and Codex have per-project memory, butwhen you switch between projects, the agent starts from scratch. "How was that API project going?" — it has no idea.

Hive Memory solves this by acting as a central memory layer. The agent automatically stores decisions, learnings, and session progress, then restores context when you come back.

Key points: - 14 MCP tools: project management, memory store/recall, session save, group organization - Works with Claude Code, Claude Desktop, Cursor, Codex — any MCP client - Auto-discovers projects in a directory and onboards them - Group system for organizing related projects with shared guides - Fully local (stored in ~/.cortex/ as JSON/Markdown). No cloud, no accounts - npm install -g hive-memory

The typical workflow: open a project → agent loads last session context → work → agent saves progress → switch to another project → repeat. Days later, come back and the agent knows exactly where you left off.

Built with TypeScript + MCP SDK. Optional Rust native module for semantic search.