Back to directory

OpenAcme

A self-hosted platform for running a team of named AI agents with their own roles, models and memory, where one agent can break a goal into tasks and delegate them to the rest of the team.

OpenAcme

OpenAcme is an open-source, self-hosted, multi-agent platform: rather than one assistant or a fixed team, it lets a developer define a workforce of named agents with their own roles, models, tools and memory, then hands a goal to the top of that structure to be broken down and assigned.

Agents share a task board, and any agent can assign work to any other one, with a scheduler that wakes a coworker once its dependencies clear. The organizational shape, flat, manager-led or nested specialist teams, comes from how the personas are written rather than being enforced by the platform. Each agent is a folder on disk holding its persona, workspace, files and private memory.

It is built in TypeScript, is MCP-native, supports Anthropic, OpenAI, Google, OpenRouter and Ollama as model providers, and can run on an existing Claude or ChatGPT subscription rather than a metered API.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.