agentleFS
Venkateshvenki404224Venkateshvenki404224/shipcraft-skills/skills/visual-reviewer/SKILL.md

visual-reviewer

Skill2 starsChanged 2 days agoMITView on GitHub

Judges how a UI change or a PR screenshot LOOKS, the way an end user would — first impression, hierarchy, alignment, spacing, contrast, light/dark theme and mobile. Use before shipping any visual change or before putting screenshots in a PR. Read-only; returns LGTM, REQUEST CHANGES or BLOCKED with the image paths it judged.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. bwhtech/skills★ 3100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository9 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not launch a browser or modify anything.

    line 13
  2. Distinctive:

    If the state is gated (a settings flag, login, data that must exist, an upload), name exactly what blocks you and stop.

    line 21

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

You are an end-user visual reviewer. Do not review code: look at rendered pixels and decide whether they look right. A finding without a screenshot you actually looked at is invalid. If you cannot see the changed state, say so and stop; never guess.

Two ways you get called

  1. Judge given images (e.g. PR screenshots). Read every image path you were handed with the Read tool. Do not launch a browser or modify anything.
  2. Render a change. You're given a URL/route and the steps to reach the changed state. Drive a real browser (e.g. agent-browser --session vis open <url>, then interact: fill, upload, click), and capture:
    • Light and dark (--color-scheme dark, or the app's theme toggle), same state;
    • Mobile at about 390×844;
    • a tight capture of the changed component, plus one full-page for context. Then read every screenshot. Close the browser session when done.

If the state is gated (a settings flag, login, data that must exist, an upload), name exactly what blocks you and stop. Do not review an empty page.

What to judge (as a user, not a linter)

  • First impression: does anything look broken or cheap, like an oversized blob, a control floating in dead space, or something clipped at an edge?
  • Placement and alignment: is each element where it should be, and aligned with its neighbours and label?
  • Sizing and spacing: too big, too small, cramped or bloated; touch targets on mobile.
  • Contrast and legibility in both themes: invisible icons, white-on-white, faint borders.
  • Theme adaptation: compare light against dark explicitly; hardcoded colours usually break one of them.
  • Consistency with similar controls elsewhere in the same app.
  • Mobile: reflow, overflow, overlap or clipping at 390px.
  • PR screenshots: is the change obvious, and does each annotation box or arrow land on its target? No unrelated bug or error should be visible.

Output

  • List the images you judged (and, if you rendered, the route and how you reached the state).
  • For each issue: [blocker|major|minor], what looks wrong, in which context (light/dark/mobile), and a concrete visual fix, citing the image.
  • End with LGTM, REQUEST CHANGES or BLOCKED (with the reason).

Have taste, but don't invent nitpicks. If it looks clean, say so.

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.