Transform a raw bug report into a structured triage summary with severity, reproduction steps, and priority recommendation.
Convert a raw bug report into a prioritized triage summary that includes severity rating, clear reproduction steps, expected vs. actual behavior, affected scope, and a recommended next action.
Invoke when a new bug report is submitted and needs to be structured, prioritized, and assigned before entering the development queue.
Analyze the raw bug report below and extract all available signal — user description, environment details, error messages, screenshots mentioned, and any logs: $bug_report
Analyze the raw bug report below and extract all available signal — user description, environment details, error messages, screenshots mentioned, and any logs:
$bug_reportAssign a severity level (Critical / High / Medium / Low) based on user impact, data risk, and frequency. Justify the rating in one sentence referencing specific details from the report.
Write clear, numbered reproduction steps starting from a clean state. Each step should be a single atomic action. Flag any steps that are ambiguous or missing from the original report.
State the expected behavior and the actual behavior observed, each in one concise sentence. Then summarize the affected scope — which users, platforms, features, or data are impacted.
Identify any likely root cause hypotheses based on the report (e.g. race condition, null pointer, config error). List 1–3 hypotheses ranked by plausibility, each with a brief rationale.
Produce a recommended next action: assign to a team or role, suggest an immediate workaround if applicable, and flag if escalation or a hotfix is warranted based on severity.
Return a structured triage summary with clearly labeled sections: Severity (with justification), Reproduction Steps (numbered list), Expected vs. Actual Behavior, Affected Scope, Root Cause Hypotheses (ranked list), and Recommended Next Action. Use plain language suitable for both engineers and non-technical stakeholders.
Copy this skill into your library to inject it into Claude, ChatGPT, and Gemini — or install your whole library as / slash commands in Claude Code and Cowork.
Get started free →Related skills
Break a large user story into small, independently shippable stories that each deliver real user value.
Draft a blameless postmortem from an incident timeline and impact summary, following SRE best practices.
Review a pull request diff for correctness, security, and maintainability with a structured checklist.