RUC team releases CoworkOS, a long-horizon agent harness separating strategy from execution
RUC · hf · 2026-11-03
A RUC-led team releases CoworkOS, a general-purpose harness for long-horizon agent work that separates task strategy from execution management: the model decides actions, delegation, and acceptance while the runtime maintains state and feedback.
- Three-layer architecture: direct execution or sync/async delegation to subagents under stable responsibilities
- Human-in-the-loop: humans and agents inspect and revise the same evolving documents, slides, CAD models, and Blender scenes
- Continuity: hierarchical workspaces and persistent semantic memory keep files, decisions, and job logs shareable and resumable after interruption
- Sustained automated post-training improves model performance on real workloads, validated across 12 benchmarks
More from coding & agent
- LetterBox adds granular API key permissions, agents can create Drafts over MCP — rchardkovacs · 2026-11-03
- Tsinghua study: continual harness self-evolution causes severe catastrophic forgetting in coding agents — Shangziqi Zhao · 2026-11-03
- Vantage: open-source Windows app lets AI agents drive your real desktop, 80-step task in ~8 min — AKGAMING1234 · 2026-11-03
- Reward hacking risks when agents optimize inference, new article warns — AccBalanced · 2026-11-03
- Build a Media Agent workshop: CPU-based agent calling GPU tools on demand — Al_Grigor · 2026-11-03
- Factory rebuilt its Droid coding agent to run inside customers' own clouds, plans to open source the compute layer — AI Engineer · 2026-11-03