# Verify Fix

> Use when a minimal fix has been applied and the original failure must be proven gone: running the regression validation target and observing it pass where it previously failed. This is the verify_fix action of the debugging stretch — one Agent-tool dispatch with subagent_type qa-engineer — the before/after proof, distinct from the test_evidence_review action (review-evidence skill, the later, broader verification pass over the whole change's evidence).

- Skill: `bhangeef16/verify-fix` (Agent Skill)
- Install (CLI): `npx skillmds@latest add bhangeef16/verify-fix`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bhangeef16/verify-fix/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: BhangeeF16 (https://skillmd.com/u/bhangeef16)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/bhangeef16/verify-fix

---


# Verify Fix

Prove the failure is gone by running what used to fail. The same reproduction that demonstrated the failure is the strongest possible proof of the fix.

## Dispatch

One synchronous `Agent`-tool call, `subagent_type: qa-engineer`.

## Required inputs

- The minimal fix and its regression validation target.

## Required outputs (all mandatory)

- Passing regression validation — the previously failing command/test run again now, observed passing, with the output recorded; plus confirmation nothing adjacent that was passing now fails.

## Rules

- Run the actual reproduction from reproduce-failure wherever possible — a different, kinder test passing is weaker evidence than the original failure not recurring.

## Result

Report each required output produced (with its location), the result (pass / pass_with_risks / blocked / failure), and blocking questions as an explicit list (empty list stated explicitly).