一键导入
trace
Evidence-driven debugging with parallel hypothesis testing
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。
菜单
Evidence-driven debugging with parallel hypothesis testing
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
| name | trace |
| description | Evidence-driven debugging with parallel hypothesis testing |
| argument-hint | <bug or error description> |
| level | 3 |
<Use_When>
<Do_Not_Use_When>
<Debugging_Protocol>
List all possible explanations for the bug. Example: "Login fails"
For each hypothesis, determine what evidence would confirm or refute it. Then gather that evidence by reading the relevant code.
The hypothesis with the most supporting evidence and no refutation is the root cause.
Implement the minimal fix that addresses the root cause.
Confirm the original symptom is resolved. </Debugging_Protocol>
<Output_Format>
## Bug Trace Report
### Evidence
- [Collected error messages, logs, context]
### Hypotheses
| Hypothesis | Evidence For | Evidence Against |
|-----------|--------------|-----------------|
| A: ... | ... | ... |
| B: ... | ... | ... |
### Root Cause
[Identified root cause with explanation]
### Fix
[Code change]
### Verification
[Test showing fix works]
</Output_Format>
User: "trace: API returns 500 only on production, works fine locally" Why good: Specific symptom, environment difference hints at config issue User: "trace: this code doesn't work" Why bad: No evidence, no symptom descriptionFull autonomous execution from idea to working code
Production-grade Docker configurations (Dockerfile, Compose, best practices)
Persistence mode — doesn't stop until the task is verified complete
Multi-agent coordinated execution
Autonomous QA cycling until all tests pass
Maximum parallel throughput — execute independent tasks simultaneously