Paper introduces "harness-level forgetting": agent prompt/memory updates can break reliable behavior
omarsar0 · x · 2026-08-21
A paper on continual learning for agent harnesses. Modern agents accumulate experience not in weights but in the harness — prompts, memories, tools, skills, and routing rules. Updating any harness component can break previously reliable behavior with the model untouched, which the paper names harness-level forgetting and provides a way to measure. Its Guarded Harness Evolution separates proposing an update from committing it: a Continual Optimizer drafts candidate harness changes from post-execution feedback, while a Continual Evaluator commits only after checking current improvement, historical retention, and validity. Relative gains exceed 10% (post truncated).
More from coding & agent
- User criticizes Cursor for opaque sub-agent model configuration — brandon_galang · 2026-08-21
- AI-native law firm Vector Legal raises seed round led by Base10 — LangChain · 2026-08-21
- Arena: Where's the line between resourceful agents and reward hacking? — arena · 2026-08-21
- Arena Dialogues: The Line Between Resourceful Agents and Reward Hacking — arena · 2026-08-21
- The hard part of AI agents isn't prompting, it's being the director — OGMYT · 2026-08-21
- CodeSurf-CLI Released: A Hackable TUI with Hot-Reloading Extensions — jasonkneen · 2026-08-21