Google open-sources Harness²: frozen weights, up to 11.5-point agent gains
lmoroney · x · 2026-10-19
Google Research open-sourced Harness², a framework that improves agents without touching model weights — instead optimizing prompts, guardrails, memory, tool guidance, subagents, skills, scripts and plugins. - For each task it proposes harness edits, probes them on that same task, and composes a refined harness from run differences, carrying lessons forward to future tasks. - Reported gains: up to 11.5 points on WorkBuddy-Bench and 4.5 on JobBench for GLM-5 and Gemini 3.5 Flash. - Apache 2.0 code, with a local demo requiring no model credentials. Suggested workflow: run the demo first, then check whether its saved edits match what you'd have done by hand.
More from coding & agent
- Qwen Code Desktop v0.25.3 Ships With Batch of Managed-Agent Fixes — github-actions[bot] · 2026-10-19
- Teknium to Turn Hermes Agent Into a DIY Game Dev Studio — Teknium · 2026-10-19
- Agent Control Plane explained: John Savill's quick overview clears up common confusion — JohnSavill · 2026-10-19
- Microsoft's ASSERT and ACS: blocking AI agents from accessing the wrong account — usamawahabkhan · 2026-10-19
- Exa Snapshot lets agents search the web as it looked on any past date — AI Engineer · 2026-10-19
- Anthropic Researcher: Claude Wasn't Worth Using at 80% Accuracy, Now Writes Most Code at 99% — victor_explore · 2026-10-19