jarvis-bug-triage-01_skill

Overview

This skill triages production bugs quickly by isolating reproducible steps and prioritizing risk-first fixes. It produces measurable success criteria, checkpointed evidence, and implementation-ready outputs. Use it to convert ad-hoc incidents into repeatable, low-risk development work with clear owners and timelines.

How this skill works

Provide the desired outcome, deadline, constraints, and available artifacts (logs, screenshots, code). The skill defines success/failure criteria, builds the smallest valid execution path, and runs checkpointed investigation steps while recording evidence. It applies quality gates, ranks findings by impact, and returns a concise action plan with exact checklist items, rollback recommendations, and owners.

When to use it

• A production bug needs a repeatable repro and an actionable fix plan.
• You must deliver a fix with constrained time, tools, or risk tolerance.
• Engineering handoff requires clear acceptance checks and owners.
• You need reversible or low-risk changes prioritized first.
• Turning messy notes or incident reports into a development-ready plan.

Best practices

• Define measurable success and failure criteria upfront.
• Start with the smallest-valid execution path to confirm repro.
• Record evidence at each checkpoint (logs, commands, screenshots).
• Prefer reversible changes and include explicit fallback plans.
• Document assumptions, tradeoffs, owners, and ETAs for every step.

Example use cases

• Convert incident notes and logs into a step-by-step reproduction and fix plan.
• Produce a risk-first mitigation sequence when a bug affects production.
• Create a sprint-ready task list with acceptance checks for a flaky feature.
• Generate rollback-safe commands and checklists for hotfix deployment.
• Outline next actions and owners after a postmortem or outage.