OpenAcme
VerifiedLocal-first platform for teams of named AI agents that delegate tasks autonomously.
What is OpenAcme?
OpenAcme is a local-first, open-source platform that lets users assemble and run a workforce of distinct AI agents. Every agent is defined with a clear role, persona, tools, and memory, plus its own MCP server instance.
Agents collaborate by delegating tasks directly to coworkers. A built-in scheduler monitors dependencies and wakes the appropriate agents when their inputs become available, enabling fully autonomous workflow progression without external orchestration.
It is designed for developers and teams who want private, on-device multi-agent systems that can handle complex projects through internal task hand-offs.
Capabilities
What you can build with OpenAcme
Software project coordination
Multiple agents handle planning, coding, testing, and documentation by passing subtasks among themselves until the full feature is complete.
Research report generation
Specialized agents gather information, analyze findings, and compile sections, with the scheduler ensuring each step waits for prior results.
Content production pipeline
Agents focused on research, drafting, editing, and formatting delegate work sequentially while keeping all data on the local machine.
Install OpenAcme
npm install -g @openacme/cli- 1Clone the OpenAcme repository from its public source.
- 2Install required dependencies using the project's setup instructions.
- 3Define agent roles, personas, tools, and MCP servers in the configuration files.
- 4Launch the scheduler to enable automatic task delegation and wake-ups.
- 5Monitor agent activity through the local dashboard or logs.
Works with
OpenAcme: pros & cons
Pros
- +Fully local execution keeps all data and models private
- +Open-source code allows inspection and customization
- +Self-organizing delegation reduces need for manual workflow scripting
- +Per-agent MCP servers provide isolated tool access
Cons
- –Requires local hardware capable of running multiple agents simultaneously
- –Initial setup of roles and MCP servers can be time-consuming
- –No built-in cloud scaling or remote collaboration features
Frequently asked questions
No, the platform is designed to run completely locally once models and tools are available on disk.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…