菜单
Use when something is broken or failing and the cause must be found before changing it. Owns failure diagnosis only.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when the direct message or commitment handoff needs the right register. Owns tone and candor only.
Use when agency, obligation, and fear are tangled. Owns ownership sorting only.
Use to extend a live idea with one useful angle. Owns additive thinking only, not mode selection or action forcing.
Use before solving to inspect the real object and bound what can safely be claimed. Owns pre-solution contact only, not diagnosis or verification.
Use when a canned frame would obscure the real object. Owns rejection of default scripts only.
Use after a claim or change needs proof before it is trusted. Owns post-claim and post-change verification only.
| name | diagnose-before-patching |
| description | Use when something is broken or failing and the cause must be found before changing it. Owns failure diagnosis only. |
| category | core |
| source_group | hermes-operational |
| compatibility | opencode |
Load this only when something is broken, failing, regressing, contradictory, or behaving differently from the stated expectation. The skill owns cause-finding before repair.
Use it for build failures, runtime errors, test regressions, bad outputs, broken prompts, unreliable workflows, or operational symptoms where a patch would otherwise be a guess.
Do not use this for ordinary pre-solution inspection, stylistic refinement, packaging a decision, simplifying a design, or proving a completed fix. If there is no failure state or symptom, this skill is too heavy.
Return the failure, evidence, likely cause, minimal patch, and the targeted check that would prove the cause was addressed.
