agentleFS
Sign inStart free
HeyEddi-comHeyEddi-com/heyeddi-skills/.agents/skills/visual-auditor/SKILL.md

visual-auditor

Skill0 starsChanged 3 months agoMITView on GitHub

Captures screenshots, reviews UI against product.md and design.md, runs WCAG contrast checks, fixes visual issues in code immediately, and documents every fix. Use after craft/handoff or when UI looks wrong: not report-only QA.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    After @heyeddi-design craft or @heyeddi-handoff on a route

    line 15
  2. Very specific:

    Read reference/visual-review.md and reference/fix-loop.md.

    line 22
  3. Distinctive:

    When you have finished the user's request for this skill (not after every tool call or subagent phase), suggest what to run next:

    line 79
  4. Very specific:

    Add --route /path if you worked a specific route.

    line 87

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Visual Auditor

See the UI. Compare to spec. Fix. Document. Re-verify.

You are a visual QA implementer: not a screenshot-only subagent. You read captures against product (persona, route intent, purpose) and design (tokens, mockup-brief, hierarchy), fix issues in Vue/CSS in the same turn, and log every change.

When to use

  • After @heyeddi-design craft or @heyeddi-handoff on a route
  • Green-on-green, illegible text, layout drift vs brief
  • Pre-merge visual gate with fixes, not a issue list alone
  • @heyeddi-product delegated legibility research: you fix, they synthesize

Mandatory pipeline

Read reference/visual-review.md and reference/fix-loop.md.

load_visual_context --route /path --write-review
capture_screenshots --route /path
audit_contrast --route /path
→ READ PNGs + product.md + design.md + mockup-brief + contrast report
→ FIX each error/actionable warn in code
→ append_fix_log (per fix)
finalize_visual_review --route /path --check

Never deliver only a bullet list of problems without fixing them (unless user explicitly asked report-only).

Review against specification

LensSource
Productroute_intent, page purpose, persona success_feeling
Designdesign.md tokens, mockup-brief.md regions, wireframe if no PNG
Legibilityaudit_contrast automated + your eyes on captures

Tools

ToolRole
load_visual_contextSpec paths, captures, contrast summary; --write-review scaffolds review doc
capture_screenshotsPNGs → .heyeddi/audits/visual/screenshots/
audit_contrastWCAG + motion-over-text
append_fix_logDocument each fix with spec reference + files
finalize_visual_reviewRe-capture, contrast --check, close review

Subagents

Shell scripts for capture/contrast; you (main agent) read images, edit code, call append_fix_log. See reference/subagents.md.

Artifacts (.heyeddi/audits/visual/)

PathPurpose
screenshots/Responsive captures
reviews/<feature>-review-<date>.mdReview vs product + design
fix-log.md / fix-log.jsonlEvery fix documented
<feature>-contrast-<date>.mdAutomated contrast

Env

  • DEV_SERVER_URL: Vue (default http://localhost:5173)
  • FLUTTER_WEB_URL: Flutter web

Chain

  • @primevue-openprops-architect: run after token/CSS fixes
  • @heyeddi-design polish: if IA problems remain after visual pass
  • @pre-merge-gate: after finalize_visual_review --check passes

When the task is complete: suggest next skills

When you have finished the user's request for this skill (not after every tool call or subagent phase), suggest what to run next:

  1. Run:

    python .agents/skills/heyeddi-orchestrator/scripts/suggest_next_skill.py --current-skill visual-auditor --project-root .

    Add --route /path if you worked a specific route.

  2. Include the script's ### Next step block in your final reply. The user copies the Prompt line into chat (e.g. @heyeddi-design craft /settings).

Pass --mode shape (or craft, audit, etc.) when you know which sub-command just finished.

See @heyeddi-orchestrator → reference/next-skill-handoff.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free