Microsoft’s OpenForgeRL trains agents inside the same harnesses they use in production
dair_ai · x · 2026-07-25
Microsoft researchers and collaborators present OpenForgeRL, a framework for training agents inside the same harnesses they run in at deployment time.
- Modern agents live inside complex harnesses such as Claude Code, Codex, and OpenClaw, but most RL training happens in simplified environments that do not match production.
- OpenForgeRL uses a lightweight proxy to serve harness model calls while recording them as training data for a standard RL stack, plus a Kubernetes orchestrator that runs each rollout in its own container.
- This lets teams train directly in the real harness at scale, reducing train-deploy mismatch.
- Reported results include 72.3 on WebVoyager, 63.0 on Online-Mind2Web, and 37.7 on OSWorld-Verified, outperforming open baselines of similar size and in some cases matching much larger models.
- The paper also finds that harness choice itself is a major variable: some harnesses are much harder to learn, RL improves self-verification and multi-step completion, but error recovery remains weak.
Related event: Microsoft Open-Sources OpenForgeRL for End-to-End Agent Training(4 posts)→
More from coding & agent
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27
- Context layout is a cost decision: one memory edit can bust your entire agent prompt cache — dscape · 2026-10-27
- Alchemy adds effect/Workflow support across Durable Objects, Workflows and Lambda Durable Functions — samgoodwin89 · 2026-10-27
- Alchemy's Lambda Web Functions run plain Node web servers on Lambda, no OpenNext needed — samgoodwin89 · 2026-10-27
- Alchemy Domain brings domain-driven design and story testing to Effect backends — samgoodwin89 · 2026-10-27
- Alchemy makes all ~2,850 AWS bindings work from any cloud, not just S3/Lambda/DynamoDB — samgoodwin89 · 2026-10-27