agentleFS
ANT1OFFANT1OFF/gram/.agents/skills/design-brief/SKILL.md

design-brief

Skill1 starsChanged yesterdayMITView on GitHub

Decide whether a task needs a mock before code, and if so write its brief and pick directions. Use when a task changes what a user sees (a screen, a flow, a control, a page) or the user asks to design, mock or sketch something first.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use brief-to-tasks, design-flow and 7 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • julianoczkowskiSkill

    brief-to-tasks

    Break a design brief into an ordered checklist of independently buildable tasks using vertical slices. Saves as a markdown checklist. Use when user wants to break down work, create tasks from a brief, plan implementation order, or mentions "tasks" or "breakdown".

    Also in 7 projects that use this

  • julianoczkowskiSkill

    design-flow

    Run the full design-to-build workflow as a guided sequence. Orchestrates all designer skills in order, from grilling through review. Use when user wants to go through the complete design process, start a project from scratch, run the full flow, or mentions "design flow" or "full workflow".

    Also in 6 projects that use this

  • 45ckSkill

    information-architecture

    Define the information architecture for a prototype -- sitemap, navigation model, grouping logic, labeling, and action placement.

    Also in 8 projects that use this

  • ilikescienceSkill

    design-tokens

    Expert in design tokens using the DTCG specification. Use this skill when users ask about design tokens, DTCG format, token validation, formatting, transformation, color spaces (sRGB, Display P3, OKLCH), references and aliasing, resolvers, theming with modifiers/contexts, multi-platform design systems, accessibility, or working with tools like jq, jsonata, and terrazzo. Helps with token file creation, resolver configuration, structure, naming conventions, and best practices.

    Also in 8 projects that use this

  • launchapp-devSkill

    color-system

    Build a comprehensive color system with palette generation, semantic mapping, and accessibility compliance.

    Also in 3 projects that use this

  • dillonluiSkill

    design-review

    Review a view or flow for UX and UI issues. Provides severity-ranked, actionable findings grounded in established design heuristics. Use when the user asks for design feedback, UX review, or UI critique.

    Also in 7 projects that use this

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 19 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When unsure, say which way you went and why in one line, and go.

    line 32
  2. Distinctive:

    Read before asking. A mock built on guesses gets thrown away.

    line 37
  3. Distinctive:

    Ask the user only when two or more decisions that change the result are open and the code and notes can't settle them.

    line 47
  4. Very specific:

    When the layout itself is the open question, the first version is two to four genuinely different low-fidelity directions on one page …

    line 55
  5. Distinctive:

    When the shape is already settled (an existing pattern, or the user described it), go straight to one hi-fi screen or flow.

    line 64

“Common” is measured: kept by at least 6 in 10 of the 19 edited versions of this file we could read. “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

Design brief

The first step of the mock-first loop: decide whether to mock at all, gather what the mock needs, and settle the open decisions before anything is drawn. The next steps are mock-build and mock-review; the whole loop and how it fits a workspace's orchestrator is in docs/orchestration/playbooks/mock-first.md when the repository has it.

1. Triage: does this need a mock?

Mock when the change has a visible shape someone could disagree with:

  • a new screen, view, panel or page;
  • a flow of more than one step, or a new state (empty, error, loading, permission) that a user will meet;
  • a control or layout change whose look is the point (a new menu, card, toolbar, dialog);
  • the user asked to see it first, or the brief leaves the look open.

Skip the mock and build directly when:

  • the change is a fix, a refactor, copy, a dependency or anything without UI;
  • it reuses an existing pattern exactly (another row in an existing list, another setting in an existing form);
  • the user already gave a picture, a mock or exact instructions;
  • it is a one-line visual fix whose result is obvious from the description.

When unsure, say which way you went and why in one line, and go. A skipped mock costs nothing to add later; an unwanted one costs a review round.

2. Ground it

Read before asking. A mock built on guesses gets thrown away.

  • The workspace's design notes: DESIGN.md at the root if present, then the theme the kit's tokens come from (docs/design/mocks/tokens.config.json names it).
  • The real code of the screens this touches, and the components it would use. Hi-fi work starts from what exists: the mock should look like the product, not like a new product.
  • Earlier mocks for the same area in docs/design/mocks/, and their decisions.md.

3. Ask once, only if it matters

Ask the user only when two or more decisions that change the result are open and the code and notes can't settle them. Then ask once: at most four questions, each answerable in a word, each with your default marked. Otherwise pick the reasonable default, write it into the brief as an assumption, and go. Never ask about things a mock answers better by showing (colours, spacing, which layout looks right): those become directions or tweaks.

4. Directions first when the shape is open

When the layout itself is the open question, the first version is two to four genuinely different low-fidelity directions on one page (template options), not one polished guess. Different means a different structure (inline vs a side panel vs a separate view), not the same layout in two colours. Give each a stable name ("A Inline", "B Panel") that the review can pick by, and use it as the board's name, so its id, its screenshots in shots/ and the question's options all match. The trade-offs go in your reply, not on the boards. The question that picks one carries the pictures: each option's screenshot and board, and the mock (mock-build, step 6), never a path in the context.

When the shape is already settled (an existing pattern, or the user described it), go straight to one hi-fi screen or flow.

5. Write the brief

Scaffold the mock folder (see mock-build), then fill its brief.md with the Write tool. Keep it under a screen:

  • Problem: what the user can't do or finds hard today, in their words.
  • Who and where: who meets this and from which screen.
  • Must show: the states and facts the mock has to cover (real names, real numbers from the code or the user; anything unknown stays a bracketed gap such as [YOUR PRICE]).
  • Out of scope: what this mock deliberately leaves alone.
  • Open questions: the assumptions you made, each with the default you picked.

Then hand on to mock-build with the template the triage chose.

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.