Hivemind
Hivemind is a Python-first runtime layer that gives multi-agent applications persistent run history, ordered step and tool-call records, and streamed execution events, without replacing orchestration frameworks such as LangGraph, AutoGen or CrewAI.
Hivemind provides runtime infrastructure for multi-agent applications rather than an orchestration framework of its own. Agents are invoked as runs, each run produces ordered steps and messages, tool calls are recorded, and every state change can be streamed to a client, giving frameworks such as LangGraph, AutoGen or CrewAI a consistent execution model to build a production service on top of.
Its default adapter uses LangGraph, with an official PydanticAI integration shipped as a plugin and room to register adapters for AutoGen, CrewAI or internal frameworks without changing the API or database schema. Run, Step, Message, ToolCall and Checkpoint are stored as first-class database entities, run outcomes are pushed to configured webhook URLs, and the stack pairs a Java Spring Boot API with Python workers, SQLAlchemy and Redis.