pf-hotfix_skill

Overview

This skill codifies a fast emergency bug-fix workflow for the PF DEV monorepo. It guides engineers through reproducing the issue, creating a hotfix branch, making minimal changes, running quick tests, creating a PR, and verifying deployment. The flow emphasizes speed, safety, and minimal surface-area changes to reduce risk during urgent fixes.

How this skill works

The skill inspects the problem by collecting reproduction steps, error messages, and recent changes, then recommends quick git commands to locate suspect commits and files. It prescribes a branch naming convention, strict rules for minimal edits, and a short test checklist (type check, lint, build, local repro). Finally, it provides PR templates, merge guidance for emergencies, deployment checks, and rollback steps if needed.

When to use it

• A production crash or severe bug requiring immediate remediation
• A regression introduced by a recent deploy that blocks users
• Customer-impacting defects with tight SLA for resolution
• Urgent bug fixes that must not include unrelated refactors

Best practices

• Reproduce and root-cause the bug within ~10 minutes before coding
• Create a dedicated fix branch from main (fix/ISSUE-number-short-desc)
• Change only what directly fixes the bug; avoid refactors or renames
• Run type checks, linter, and build locally before opening PR
• Use a concise PR body describing bug, cause, fix, and tests
• Verify deployment and confirm no side effects; revert if necessary

Example use cases

• Login page crashes when users array is empty — add a null check and deploy
• API client failing due to recent header change — patch header handling and validate
• Frontend UI throws runtime error after last deploy — create hotfix branch, fix, and merge
• Critical production bug reported by SLA monitoring that must be resolved immediately