Agensh is a scalable, research-oriented multi-agent harness that enables large teams of AI coding agents to collaborate without a central orchestrator.
Follow Agensh to get an email when its listing changes, such as new pricing or features. One digest a day at most, and you can unsubscribe anytime.
It facilitates asynchronous cooperation by using a shared Gitea workspace, a messaging interface, and a shared context mechanism.
- 01
Decentralized cooperation loop without a central orchestrator
- 02
Shared workspace, messaging interface, and context tracking
- 03
Support for multiple agent backends including Copilot, Codex, Claude Code, and Pi
- 04
Scalable architecture capable of coordinating up to 1,024 agents
- 05
Single-node team deployment via Docker Compose
Open Source
Free and open-source under the MIT license; users are responsible for the costs of LLM inference for multiple concurrent agents and the required Linux infrastructure.
Best for
Researchers and platform engineers experimenting with multi-agent coordination patterns or attempting to reproduce the authors' ProgramBench experiments.
Not ideal for
Developers seeking a turnkey, daily-use coding assistant; the tool requires significant Linux infrastructure (including Docker Engine, systemd, and cgroup v2) and manages complex, long-running agent tasks.
A valuable, open-source research harness for studying multi-agent scaling and peer coordination, though it is intended for experimental use rather than as a production-ready application.
On 2026-10-29, support was added for Copilot, Codex, Claude Code, and Pi as single-agent backends. The project was initially open-sourced on 2026-10-24.
As of
Is Agensh free?+
Yes - Agensh is Open Source. Free and open-source under the MIT license; users are responsible for the costs of LLM inference for multiple concurrent agents and the required Linux infrastructure.
Is Agensh open source?+
Yes - Agensh is open source. Free and open-source under the MIT license; users are responsible for the costs of LLM inference for multiple concurrent agents and the required Linux infrastructure.
Who is Agensh best for?+
Researchers and platform engineers experimenting with multi-agent coordination patterns or attempting to reproduce the authors' ProgramBench experiments.
Who is Agensh not ideal for?+
Developers seeking a turnkey, daily-use coding assistant; the tool requires significant Linux infrastructure (including Docker Engine, systemd, and cgroup v2) and manages complex, long-running agent tasks.
What are the best Agensh alternatives?+
The closest Agensh alternatives on ai.dosa.dev are Jules by Google, OpenHands, SWE-agent - all listed under Autonomous & Async Agents.
Who makes Agensh?+
Agensh is developed by Microsoft. It is listed in the Autonomous & Async Agents category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.