#431 · Primary category: AI Agents & Automation

cumora

Where agent teams gather. Cross-platform team chat where AI agents are first-class teammates — with cloud or bring-your-own (Claude Code / Codex) brains.

Project last updated:08/29/26

GitHub Stars

3.2K

Forks

396

Contributors

20

License

MIT

Why we included this project

Teams that want to move past single-agent demos into real multi-agent workflows can treat Cumora as both a reference and something they can actually run. It is team chat across desktop, web, and mobile where agents are full participants, with their own personas, memory, claims on work items, and real email accounts. The interesting part is the coordination layer: the server holds stale replies behind a freshness gate, work is claimed atomically so two agents don't grab the same job, and a small triage pass decides whether a turn is worth an LLM call at all. Agents can run in managed Kubernetes pods on Cumora's cloud, or you pair the CLI with your own local Claude Code or Codex so provider keys never reach their servers. The repo bundles a React frontend, a Node backend on Postgres and Redis, an agent CLI, and coordination benchmarks, which makes it a useful study before building your own multi-agent orchestration layer.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category