Install
$ agentstack add skill-tommylower-cortex-studio-audit ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
studio-audit
Use this when the user feels "done" and wants the work judged like a senior designer would judge it before handoff. This is an umbrella review skill, not a new visual system. It composes the existing Cortex design review and craft skills, then synthesizes one priority-ordered report.
law
Read [references/studio-law.md](references/studio-law.md) before judging. Read [references/report-template.md](references/report-template.md) before writing the final report.
Default to audit-only. Do not edit code, run formatters, or apply fixes unless the user explicitly asks for fixes. If a delegated skill offers to fix safe small issues, keep it in report-only mode unless the invocation includes fix, apply, or equivalent explicit permission.
workflow
- scope the artifact. Identify whether the target is a component, page,
flow, whole app, screenshot, URL, or branch. If scope is missing, audit the current visible app or current branch and state that assumption.
- load project truth. Read local design docs, existing component
conventions, recent changed files, and any asbuilt package if one is present. Do not require asbuilt; treat it as extra project-specific evidence.
- static ship check. Use
preflightfor code-level accessibility,
visual consistency, responsive footguns, and AI-pattern tells.
- viewport check. Use
responsive-craft previewor its snapshot workflow
when the target can run in a browser. At minimum cover mobile and desktop; note any untested viewport range.
- live experience check. Use
wip-senior-auditas the live-site engine
when the app can boot or a URL is available. It covers first impressions, navigation, hierarchy, consistency, states, trust, and conversion. If the app cannot run, mark live flow coverage as not covered.
- taste and interaction check. Use
interface-craft critiquefor
composition, hierarchy, density, user context, and pattern cohesion. Use emil-design-eng for motion, feedback, timing, and interaction feel.
- synthesize. Deduplicate findings across lenses. Order by severity,
not by the skill that found them. Every finding must name the evidence, why it matters, and a concrete fix.
output
Write the final answer in the format from [references/report-template.md](references/report-template.md).
If working inside a repo and the audit is more than a quick chat critique, also write the durable report to docs/design-audit/studio-audit.md. If screenshots were captured, store them under docs/design-audit/assets/ and reference them from the report.
Close with:
ship: no blocking issuesfix first: P0/P1 issues existreview again: the audit could not cover the live flow, mobile, or another
required surface
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: tommylower
- Source: tommylower/cortex
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.