Amazon Study Finds Stale Agent Skills Hurt After Environment Changes

Original title: New Amazon paper finds that saved agent skills can hurt once the environment changes, and checking a...
74AI Score

A new Amazon paper finds that saved agent skills such as file paths, field names and commands silently go stale after app or API changes, hurting performance. On Terminal-Bench 2, a skill that had added 4.7 points for Opus cut pass rate 11.4 points below no-skill after the change; their selective check-and-repair approach lifted Haiku's tax-form field accuracy from 30.5% to 97.5%.

Full Report

Rohan Paul (@rohanpaul_ai) · 2026-11-19 19

New Amazon paper finds that saved agent skills can hurt once the environment changes, and checking and repairing them selectively restores their value.

Agent skills store past experience, like file paths, field names, or commands, so agents can skip work later. When an app or API changes, some of that stored detail goes stale without warning.

They tested Claude Opus 4.7 and Haiku 4.5 on 4 settings, including Terminal-Bench 2 and IRS tax forms, before and after a change. Their fix replays saved checks witho1ut model calls and repairs or rebuilds only the conflicting skill content.

On Terminal-Bench 2, a skill that had added 4.7 points for Opus cut its pass rate 11.4 points below no skill after the change. Maintenance lifted Haiku's tax-form field accuracy from 30.5% with the stale skill to 97.5%.

Check saved agent skills against the current environment before reuse, instead of waiting for tasks to fail.

推文配图