agentleFS
a-lottesa-lottes/aSPARK/skills/look-and-feel/SKILL.md

look-and-feel

Skill25 starsChanged 3 months agoMITView on GitHub

Run the Designer over a spec (design risks, usability heuristics, accessibility) or over an implemented UI (screenshots, running app). Use after /story-time for UI-facing features, or any time the user wants a design critique of a spec, mockup, page or component.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 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.

  • 2

    Very specific

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

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

    Resolve the feature: If a feature name was given, use .spark/<feature-name>/.

    line 21
  2. Very specific:

    Read the active lenses from .spark/constitution.md (its *Project Profile* section).

    line 33
  3. Distinctive:

    Relay evidence requests: If the agent asks for missing evidence (screenshots, flows, viewports), get it from the user or the browser and …

    line 46

≈ “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

/look-and-feel — Specify (Designer)

You are running the design check ceremony. The Designer detects bad design before it gets planned — or critiques a UI that already exists.

Input

Optional argument: a feature name, a URL, or paths to screenshots/components.

Steps

  1. Resolve the feature. If a feature name was given, use .spark/<feature-name>/. Otherwise: if .spark/ holds exactly one feature, use it; if several, ask the user which one.
  2. Pick the mode.
    • Mode A (default): design-check the spec. Requires .spark/<feature-name>/spec.md to exist — if it doesn't, stop and point the user to /story-time.
    • Mode B: the user provided a URL, screenshots or component paths — critique the implemented UI. If a URL was given, capture evidence first (screenshots via the available browser tooling) so the agent judges what is actually rendered.
  3. Resolve active lenses. The constitution is the single source of truth. Read the active lenses from .spark/constitution.md (its Project Profile section); pass the path of every active lens whose own frontmatter phases field includes design in step 4. To decide, read each active lens's file (${CLAUDE_PLUGIN_ROOT}/lenses/<name>.md); do not work from a list of lens names given here. If there's no constitution, do not apply lenses off a guess — only give a lightweight nudge: name the likely type in one line (e.g. "app-like web-app — a ux lens would apply") and point the user to /charter to record it. No lens is switched on without a confirmed constitution entry.
  4. Delegate to the Designer. Invoke the designer agent with the mode, the feature paths, any evidence, and the paths of the design-relevant active lenses (${CLAUDE_PLUGIN_ROOT}/lenses/<name>.md). In Mode A it fills the Design Review section of the spec; in Mode B it returns a findings report.
  5. Relay evidence requests. If the agent asks for missing evidence (screenshots, flows, viewports), get it from the user or the browser and re-invoke.
  6. Present the findings by severity, each with location, violated rule and suggested fix. If the Designer raised scope questions, route them explicitly back to the PO: offer to run /story-time on the spec again.
  7. Close the gate (Mode A). With the design review filled in, walk the SPEC GATE checklist with the user. Required design changes go into the spec (stories/ACs adjusted via the PO if needed). On the user's explicit approval, set the spec status to approved.

Rules

  • Findings without location + rule + fix don't get presented — send them back to the agent.
  • Blocker design findings block the SPEC GATE like any other open question.

Handoff

  • Spec approved → /sprint-plan
  • Mode B critique after implementation → findings feed /increment (fixes) and re-check with /look-and-feel afterwards.

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.