研究提出技能粒度运行时自适应，让编程智能体失败时回退到原子步骤
研究者在做编程智能体的技能调度：任务跑宏失败就自动退回原子步骤，12 个任务路由实验里学习路由器拿到 100% 成功率，做 Agent 的可以看看。
这项研究把技能（skill）粒度当作运行时决策：智能体面对每个任务可以选择原始工具调用、宏技能，或带失败回退的宏技能。控制器依据任务结构、执行状态和历史宏失败来选择粒度，由有限状态执行器把回退处理为显式状态转移而非简单重试。在 12 个任务的路由实验中，学习到的路由器达到 100% 成功率，而手写启发式规则失败。宏到微观回退机制能恢复固定粒度方案无法解决的受控宏失败。
Nice work on using skills more effectively for coding agents.
(bookmark it)
In this work, researchers treat skill granularity as a runtime decision. For each task, the agent can use primitive tool calls, a macro skill, or a macro skill that falls back to primitive steps when it fails.
Fixed macro execution is efficient when the skill fits and brittle when it does not.
The controller picks the level from task structure, execution state, and earlier macro failures, and a finite-state executor handles the fallback as an explicit transition instead of an informal retry.
In a 12-task routing study, a learned router reaches 100% success where a hand-written heuristic fails. With macro-to-micro fallback, the agent recovers controlled macro failures that fixed granularity leaves unresolved.
https://t.co/Ze4OjNcK5S
Chat with Paper: https://t.co/yfLayeF1pT
