Skip to content
Back to skills

Demo Audit

ASecurity

Independently audit a demo's actual screen, speech, action coverage, transitions, and evidence before it is shared.

  • 5 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 7, 2026
research

Works with

  • cursor

Security analysis

A100/100

Pro scans all 5 files and shows the line behind each finding

Scanned November 7, 2026

npx -y skills add Wanfeng1028/muse --skill demo-audit --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Demo Audit?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Demo Audit
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/wanfeng1028-demo-audit/badge)](https://www.skillsdirectory.com/skills/wanfeng1028-demo-audit)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: demo-audit
description: Independently audit a demo's actual screen, speech, action coverage, transitions, and evidence before it is shared.
---

# Demo audit

Use this skill after capture or when a demo needs an independent quality verdict. Compare the actual video and audio with the approved story and current plan. The person or agent that wrote the narration cannot be the sole reviewer.

Sample the screen and speech at regular intervals and around every active action and scene boundary. Inspect the final third separately for pacing and action density. Check product truth, full-frame legibility, spoken coverage, captions, cursor continuity, transitions, accessibility, and privacy.

## Verdict

Report concrete evidence and a location for every finding. A missing recording, unavailable audio, stale approval, or unverified transition is inconclusive—not a pass. Escalate P0/P1 findings for operator review and stop publication until resolved. Do not invent a passing result when evidence is absent. A render that clears the media audit waits for review; it is not ready to publish. Mirror the audit's automated checks in a silent rehearsal so they fail before voice time is spent.

Read [the audit contract](references/audit-contract.md) for sampling, severity, bounded repair, and the final report. Before calling a video ready for a course platform or public channel, run [the delivery checks](references/delivery-checks.md) for phone-size legibility, a muted watch, loudness, caption reading limits, flashes, length, and sidecar captions, transcript, and chapters. Return findings to the earliest input that owns them; rerun the affected downstream checks after repair.

Files in this skill

  • LICENSE1.1 KB
  • SKILL.md1.7 KB
  • agents/openai.yaml218 B
  • references/audit-contract.md3.2 KB
  • references/delivery-checks.md5.7 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.