CatalogClawHub · weiyunbyte

self-improving-agent

记录错误、纠正、能力缺口与最佳实践,形成可复用的持续改进闭环。适用于:命令失败、用户纠正、外部 API/工具异常、发现更优做法、提出新能力需求,以及任务前复盘历史经验。

Pending analysisCommunity66 installs

Savant verdict: Pending analysis

Imported from the source hub; Savant's analysis is still running.

Live evaluation

Not evaluated yet. Workspaces can request a live evaluation.

Security assessment

Scan pending.

SKILL.md

This listing is enumerated but its package hasn't been fetched yet; it's queued by popularity.