# Acceptance Audit

> Use when a fully verified slice needs its structured acceptance audit before the final decision: assumptions surfaced, edge cases enumerated, a validation checklist built from the actual evidence, and residual risks stated. This is the acceptance_audit action of the acceptance stretch — one Agent-tool dispatch with subagent_type qa-engineer — the audit that feeds the delivery-lead's final accept/block/return decision, not the decision itself.

- Skill: `bhangeef16/acceptance-audit` (Agent Skill)
- Install (CLI): `npx skillmds@latest add bhangeef16/acceptance-audit`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bhangeef16/acceptance-audit/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/acceptance-audit

---


# Acceptance Audit

Audit the whole slice's evidence the way a skeptic would before signing. The audit's product is the material a defensible final decision needs — assumptions, edge cases, checklist, risks — not the decision.

## Dispatch

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

## Required inputs

- The verification evidence.
- The changed files.
- The validation output.

## Required outputs (all mandatory)

- Assumptions — everything the evidence takes for granted, stated so it can be challenged.
- Edge cases — covered and consciously-not-covered, each with its reason.
- A validation checklist mapping each acceptance criterion to the actual evidence that proves it.
- Residual risks — explicit, even when the list is empty.

## 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).