{"skill":{"slug":"complex-bug-debugging-with-ai","displayName":"complex-bug-debugging-with-ai","summary":"复杂 bug 与 AI 协作排查的元方法论。当用户报告\"诡异 / 间歇性 / 多层因素 / 重启不愈 / 多系统协作 / 已经排查很久没头绪\"的 bug 时，启用本 SKILL 的 7 阶段协作工作流：\"业务链路对齐 → 症状结构化 → 边界测试循环 → 方案摆台 → 执行验证 → 失效升级 → 闭环沉淀\"。本...","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":46,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1777255725261,"updatedAt":1777258311980},"latestVersion":{"version":"1.0.0","createdAt":1777255725261,"changelog":"Initial release: provides a rigorous, disciplined 7-stage workflow for collaborative debugging of complex bugs with AI and human operators.\n\n- Clearly defines triggers and preconditions for using this workflow (e.g. only for hard, multi-factor, not-obvious bugs).\n- Enforces model and tooling pre-check: only proceed if strong model (Opus 4.7 or similar) and all required abilities are available.\n- Introduces strict \"red line\" rules constraining both AI and user behavior to maximize joint effectiveness and avoid common failure patterns.\n- Documents a structured process (A–G stages): from business chain alignment, symptom structuring, boundary testing,方案 selection, execution/validation, failure escalation, to knowledge capture.\n- Provides copy-pasteable response templates for handling typical collaboration breakdowns.\n- Includes concise summary tables for process gates and anti-patterns to standardize practice and ensure repeatability.","license":"MIT-0"},"metadata":null,"owner":{"handle":"hgvgfgvh","userId":"s174x9jg32fvtnq226ytye9w3d83h6tm","displayName":"hgvgfgvh","image":"https://avatars.githubusercontent.com/u/150424835?v=4"},"moderation":null}