跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
开发与测试

x-safe-edit

Small, reversible edits with an explicit verification step. Use when changing code or assets where risk of breakage is high or scope must stay tight.

Fromlan · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add Fromlan/X-agent --skill x-safe-edit

技能说明


name: x-safe-edit description: Small, reversible edits with an explicit verification step. Use when changing code or assets where risk of breakage is high or scope must stay tight.

Safe edit

Discipline

  1. Scope — Touch only files needed for the agreed behaviour.
  2. Step size — Prefer one coherent change set over drive-by cleanups.
  3. Reversibility — Avoid irreversible data/format migrations unless requested.
  4. Verify — After edits, state how to check: unit test command, build, or (for Godot) reload/play via editor RPC if tools are enabled.

Stop conditions

If the change expands beyond the brief, pause and confirm with the user before continuing.