qm – Multiplayer agent harness for work
8.2 relevance
Score Breakdown
technical depth 8
novelty 9
actionability 7
community 9
strategic 7
personal 10
Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.
Multi-agent harness for work, perfect match for reader.
Summary
Qm is an open-source multi-agent harness for startups, where each employee gets a scoped workspace with isolated memory, files, and a durable sandbox, while still collaborating in Slack channels and projects. Its headless core runs a fixed tool surface including an execute command in per-scope sandboxes, and supports swapping harnesses (Pi, OpenCode, Codex, Claude Code) and models via interfaces. Security offers three postures: Strict (manual approval), Auto (AI-based content screening), and Dangerous (no restrictions), configurable at the org level.
Author
yc-software