企业官网建设流程全解析

Investigation Report: {Hypothesis Title}

【免费下载链接】agentsMulti-harness agentic plugin marketplace for Claude Code, Codex, Cursor, OpenCode, GitHub Copilot, and Google Antigravity项目地址: https://gitcode.com/GitHub_Trending/agents24/agents

Verdict: {Confirmed | Falsified | Inconclusive}

Confidence: {High (>80%) | Medium (50-80%) | Low (<50%)}

Confirming Evidence

1. src/api/users.ts:47— {发现描述}
2. src/middleware/auth.ts:23— {发现描述}

Contradicting Evidence

1. tests/api/users.test.ts:112— {反证描述}

Causal Chain (if confirmed)

1. {第一因} →
2. {中间效应} →
3. {可观测症状}

Recommended Fix

{具体代码变更及位置}

Additional Notes

{可能与其他假设相关的发现}

[parallel-debugging/SKILL.md](https://link.gitcode.com/i/358ada0af98e6a6b4c3315afd12f7de5) 还给出了证据强度分级表，这是仲裁时衡量证据质量的标尺： | 证据类型 | 强度 | 示例 | | --- | --- | --- | | Direct（直接证据） | Strong | `file.ts:42` 处的 `if (x > 0)` 应为 `if (x >= 0)` | | Correlational（相关性） | Medium | 错误率在 commit `abc123` 之后上升 | | Testimonial（口述） | Weak | "在我机器上是好的" | | Absence（缺失性） | Variable | 代码路径中未发现 null 检查 | 置信度判定标准：High（>80%）要求多份直接证据 + 清晰因果链 + 无反证；Medium（50-80%）为部分直接证据 + 合理因果链 + 轻微模糊点；Low（<50%）为以相关性证据为主、因果链不完整、存在反证。 ## Phase 5：仲裁（Arbitration） 所有 investigator 报告齐备后，主流程执行三步仲裁： **第一步：对比各 investigator 的结论**，把每个假设置于三种状态之一： - Confirmed（高置信度确认）； - Falsified（存在反证）； - Inconclusive（证据不足）。 **第二步：对确认的假设排序**，依据四项标准（与 SKILL.md 的仲裁协议一致）： 1. 置信度（High > Medium > Low）； 2. 因果链的强度； 3. 支持性证据的数量； 4. 是否存在反证。 参考文档给出的完整仲裁决策树：

All investigators reported? ├── NO → Wait for remaining reports └── YES → Count confirmed hypotheses ├── 0 confirmed │ ├── Any medium confidence? → Investigate further │ └── All low/falsified? → Generate new hypotheses ├── 1 confirmed │ └── High confidence? │ ├── YES → Declare root cause, propose fix │ └── NO → Flag as likely cause, recommend verification └── 2+ confirmed └── Are they related? ├── YES → Compound issue (multiple contributing causes) └── NO → Rank by confidence, declare highest as primary

即：无假设被确认时继续深入或重新生成假设；仅 1 个确认且高置信则宣布为根因；2 个以上确认且相互关联则判定为复合问题（多个共同作用的原因）。 **第三步：输出根因分析报告**，命令规定了固定报告模板：

Debug Report: {error description}

Root Cause (Most Likely)

Hypothesis: {description}Confidence: {High/Medium/Low}Evidence: {summary with file:line citations}Causal Chain: {step-by-step from cause to symptom}

Recommended Fix

{specific fix with code changes}

Other Hypotheses

• {hypothesis 2}: {status} — {brief evidence summary}
• {hypothesis 3}: {status} — {brief evidence summary}

【免费下载链接】agentsMulti-harness agentic plugin marketplace for Claude Code, Codex, Cursor, OpenCode, GitHub Copilot, and Google Antigravity项目地址: https://gitcode.com/GitHub_Trending/agents24/agents