Install any skill in seconds. Free to start, no credit card required.
Get Started Free →flywhlway/dtc-root-cause
远程故障码(DTC)根因分析:按 ECU 拓扑关联多码、区分首发码与伴生码、结合冻结帧定位。当用户给出 P/B/C/U 开头故障码、提到 DTC、故障灯、报文丢失、"这批车都报同一个码"时使用。不要用于 OTA 升级失败伴生的 DTC(用 ota-log-diagnosis),即使升级失败伴随 DTC。
Passed 19 of 22 test cases with the skill, and 5 of 22 without it.
and 19 more in the full report
14 fixed5 passed either way3 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
关联分析脚本输出模式为 `NO_CONFIRMED`,车辆所有故障码均处于 pending 状态未形成 confirmed。应依据 SOP 哪一节处理?
| Checked | Without skill | With skill |
|---|---|---|
| Navigates to Section 5 of the SOP document. | Missed | Passed |
The instructions your model reads when the skill loads.
多码并发时找出首发根源码,把伴生码从处置清单里剥离。 核心方法:时序 + 拓扑 + 电源域三重证据交叉,单一证据不下结论。 脚本负责关联与聚类等确定性工作;模式辨析(尤其是反例陷阱)是你的工作。
references/dtc_structure.md
确认编码含义,结合 references/root_cause_sop.md 对应模式给排查方向,
不跑脚本。但必须声明:单码无冻结帧、无伴生码上下文,
结论为方向性而非确定性。python scripts/parse_dtc.py <导出文件> -o dtc_norm.json
支持 CSV 与 JSON 两种导出格式。看 stderr:
WARN: first_seen 全部相同:TSP 平台按读取时刻覆盖了首发时间,
首发排序将降级为 occurrence 计数器排序——报告中必须声明置信度降低。WARN: first_seen 与 occurrence 均不可用:只能做拓扑分析,
报告须明确声明该局限,不要假装能排首发。python scripts/correlate_dtc.py dtc_norm.json -o correlated.json
脚本做三件确定性工作:
test cases passed, vs 5 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.