#175 · Primary category: AI Agents & Automation

qm

ai assistant harness qm

Multiplayer agent harness for work.

Project last updated:08/28/26

GitHub Stars

14.3K

Forks

1.7K

Contributors

7

License

MIT

Why we included this project

Most agents are designed like personal assistants, so getting one to work for a whole company gets complicated. QM approaches this from the other side: every person on a small team gets their own isolated workspace, and the same agent still collaborates across Slack channels and projects. The self-hosted runtime keeps per-scope memory, files, permissions, and a durable sandbox, with crons for background jobs and an admin panel to set the security posture. The agent loop is pluggable, so Pi, OpenCode, Codex, and Claude Code all drive the same backend, which avoids locking a deployment to a single vendor. If you want a company-wide agent on your own infrastructure instead of paying per seat for a hosted assistant, this is worth a close look.

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