agentleFS
bohewubohewu/agents_pipeline/skills/ui-ux-workflow/SKILL.md

ui-ux-workflow

Skill1 starsChanged 4 months agoMITView on GitHub

Create a compact concept or durable conceptual handoff for one UI surface or journey. Use for wireframes, flows, states, or prompt and preview guidance.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

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

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

    Do not produce implementation-ready layouts, component/API/data contracts, acceptance criteria, code, editable prototypes, live …

    line 15
  2. Very specific:

    use ../ui-communication-designer/SKILL.md as a compatible lens for communication-first diagnosis or microcopy

    line 17
  3. Very specific:

    use ../frontend-aesthetic-director/SKILL.md for authorized implementation after approval

    line 18
  4. Very specific:

    use $run-ux with suitable browser tooling for browser evidence and scoring

    line 19

≈ “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

UI/UX Workflow

Produce a human-reviewable concept for one bounded experience. Approval-oriented describes the artifact's purpose; it does not require renewed approval when the direction is already approved. Stay conceptual rather than producing engineering specifications or implementation.

Boundary and Pairing

This skill may produce a scoped assessment, low-fi structure, mid-fi direction, conceptual user/data/operation flows, state transitions, a provider-neutral prompt, and a thin external or read-only preview handoff.

Do not produce implementation-ready layouts, component/API/data contracts, acceptance criteria, code, editable prototypes, live integrations, or final branded design. Other skills or workflows are descriptive next steps only and require separate authorization:

  • use ../ui-communication-designer/SKILL.md as a compatible lens for communication-first diagnosis or microcopy
  • use ../frontend-aesthetic-director/SKILL.md for authorized implementation after approval
  • use $run-ux with suitable browser tooling for browser evidence and scoring

Choose the Output Depth

Compact concept

Use a compact concept when the user asks one bounded question, such as choosing a layout, clarifying a state, or sketching a short flow. Include only the sections needed to decide that question. Do not require a durable artifact, full flow set, schema, another workflow, or suggested next stage.

Identify the surface or journey, primary task, primary user, device priority, important non-happy-path state, and material content or trust constraints. Ask only when an answer would materially change the concept; otherwise label a conservative assumption.

Keep one dominant focus and primary task per screen. Use the smallest conceptual layout that supports the task, keep critical information inline, and cover relevant empty, loading, error/recovery, confirmation, and success states. Describe hierarchy, density, theme posture, progressive disclosure, and adaptation conceptually without inventing final tokens or implementation details.

Make direction specific enough to visualize: name the dominant region, reading order, relative panel emphasis, useful information density, and what should remain visually quiet. "Modern" or a style label alone is insufficient. A data-entry or comparison tool should prioritize records and controls; an editor should prioritize its canvas; an editorial page may prioritize a headline and media. These are conceptual choices, not pixel layouts or new required sections.

Use supplied references at their actual fidelity: preserve approved structure and content priority, identify transferable choices, and mark inaccessible references as uninspected. Include affected locale/long-content considerations. Do not require a moodboard, generated image, paid kit, new framework, or renewed approval for an already bounded direction.

Cover only devices explicitly requested or already required by the product's established support scope. A desktop-only concept does not acquire tablet or mobile deliverables; responsive or mobile work must describe requested adaptations clearly enough for review.

Finish when the bounded question is answered, the main action and relevant states are clear, device scope is respected, assumptions are visible, and optional follow-on work has not been treated as authorized.

Durable v1 bundle

Use a durable bundle when the user asks for a complete conceptual handoff, a versioned artifact, all flows/states, or repo-owned bundle files. Before drafting or exporting it, read ../../protocols/UI_UX_WORKFLOW.md and follow its nine-heading Markdown, five-artifact JSON, schema pairing, versioning, communication fields, template, output-path, and completion contracts. Do not produce the durable bundle from this summary alone.

The JSON is canonical when paired Markdown exists. Write repo-owned files only when the user requests files; otherwise return the handoff inline. For file output, validate against ../../protocols/schemas/ui-ux-bundle.schema.json and consult ../../protocols/examples/ui-ux-bundle.valid.json when structural guidance is needed.

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.