菜单
Use when diagnosing Eclipse SUMO, TraCI, netconvert, duarouter, route, demand, detector, traffic-light, output, teleport, insertion, seed, performance, or reproducibility failures in traffic-simulation workflows.
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。
Use when planning, coding, debugging, auditing, comparing, or writing claims for Eclipse SUMO/TraCI traffic-signal experiments, including OSM/netconvert networks, TLS/NEMA, controllers, routes, demand, detectors, outputs, baselines, metrics, reproducibility, TDD changes, and reusable field lessons.
Use when planning, coding, debugging, auditing, comparing, or writing claims for Eclipse SUMO/TraCI traffic-signal experiments, including OSM/netconvert networks, TLS/NEMA, controllers, routes, demand, detectors, outputs, baselines, metrics, reproducibility, TDD changes, and reusable field lessons.
Use when turning a confusing simulation or controller experiment result into a precise prompt, plan, or next diagnostic step. Especially relevant when comparing variants, questioning why performance changed, inferring mechanisms from logs, building fair baselines, or preserving evidence boundaries around completion, backlog, action timing, trigger logic, and metrics.
基于 SOC 职业分类
| name | debugging-helper-skill-for-eclipse-sumo |
| description | Use when diagnosing Eclipse SUMO, TraCI, netconvert, duarouter, route, demand, detector, traffic-light, output, teleport, insertion, seed, performance, or reproducibility failures in traffic-simulation workflows. |
Use this skill to debug Eclipse SUMO/TraCI failures as a closed-loop investigation. The goal is root-cause evidence, not keeping a run alive at any cost.
Loop:
failure -> observe artifacts -> classify deviation -> choose one probe -> run probe -> compare feedback -> fix, rerun, or demote
When the cause is not already proven, read only the needed references:
references/debug-sumo-traci.md.Ask for one missing artifact at a time. Prefer exact commands, SUMO logs, output files, configs, and minimal reproductions over screenshots or descriptions.
--ignore-route-errors, --ignore-errors, larger retries, --time-to-teleport -1, GUI-only inspection, or changed demand unless the run is marked diagnostic.environment-fault, input-construction-fault, demand-realization-fault, traci-protocol-fault, controller-logic-fault, output-observability-fault, stochasticity-fault, simulator-limitation-or-bug.simulation-helper-skill-for-eclipse-sumo and its field-lesson workflow after privacy/source checks.End every pass with:
root_cause:
evidence:
fix_or_next_probe:
claim_status:
residual_risk:
If the root cause is unknown, say so and keep the run out of formal evidence.