agentleFS
robritacca-dotcomrobritacca-dotcom/rift-ds/.claude/skills/new-figma-template/SKILL.md

new-figma-template

Skill9 starsChanged todayMITView on GitHub

Recreate a screen or a flow in the Rift DS Figma kit from an existing source (a template route on the site, a lab, a prototype URL, a mock image or another Figma file), built from kit instances bound to the kit's variables, never pasted as a picture. Lays it out on the kit's Templates page at web and phone sizes, chat closed and open, with each AI step of a flow as its own frame and the flow wired as a prototype. Use when asked to put a template, a prototype or a mock into Figma, or to draw a flow in the kit.

The file

We split this skill into the rules it gives an agent

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

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use when asked to put a template, a prototype, a mock or a flow into Figma.

    line 17
  2. Very specific:

    Before the first Figma call, do what the new-figma-component skill's "Before the first Figma call" section says.

    line 23
  3. Distinctive:

    A mobile web column: when the route has a narrow web layout of its own.

    line 40

≈ “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

new-figma-template

Rebuild an existing screen or flow in the Figma kit from the kit's own components. The source already exists; this skill draws it faithfully and invents nothing.

When invoked

Use when asked to put a template, a prototype, a mock or a flow into Figma. The source can be:

  • A template route (/templates/<slug> on the site, the template's code under website/src/components/templates/). The merchant sales frames on the kit's Templates page are the reference build.
  • A lab or any other route in the site, local or deployed.
  • A prototype or mock from outside: a URL, a screenshot or image, or a frame in another Figma file.

Before the first Figma call, do what the new-figma-component skill's "Before the first Figma call" section says. Its "The kit's conventions" section is the standard for every frame here.

Instructions

1. Read the source

  • A route: open it in the built-in browser at the widths you will draw (step 2), with the chat closed and open, and screenshot each. Read the template's code for its composition, its components and its pinned demo data (charts and tables are drawn from that data, never guessed from a picture). Note the theme the route runs, read off its page.
  • An image or outside prototype: read every screen. List each control and match it to a library component. A control the library has no component for is a stop: report it and ask whether it needs new-component (and then new-figma-component) first, or a drawn stand-in clearly labelled as one.

List the screens and, for a flow, the steps between them (what the person clicks or types, and what changes).

2. Plan the frames

On the kit's Templates page, one section per source, with the kit's header card at its top (the same header the component pages use). Lay out:

  • Chat closed: the screen at desktop and a large desktop width, and on a phone in a device frame with the platform's top bar and tab bar (the iOS ones, from the kit).
  • Chat open: the docked panel on web, the full-screen takeover on the phone. The responsive thresholds are the chat widget's (src/chat/queries.ts).
  • A mobile web column when the route has a narrow web layout of its own.
  • Flows: one viewport-sized frame per step, in order, left to right.

Web frames that scroll run the full length of the page with the sidebar held at viewport height; chat-open and flow frames are viewport size. Skip any row the source does not have. Give each frame a caption naming the state, linked to the source's page under SITE_URL when it has one.

3. Build from the kit

  • Instances, never pictures. Every control is an instance of the kit's component in the right variant, every surface a frame bound to the kit's variables, so the page's Theme and Mode still drive it. The only stills are what the kit already treats as stills (shaders, maps, a globe).
  • The app shell is the kit's shell for the mode the source uses (flush, floating or inset), with the sidebar and top bar from the kit.
  • Data comes from the source's pinned data, typed as the source shows it. Charts are vectors drawn from the data.
  • Build one frame, check it, then derive the rest from it (the phone and chat-open frames start as copies of the desktop one), so a fix lands once.

4. Wire the flow

For a flow, add prototype reactions: each frame's trigger (the button clicked, the chip pressed, the field submitted) navigates to the next step, with the source's transition where it has one. Set the first frame as the flow's starting point and name the flow after the source.

5. Check fidelity

Pin the source's theme on the section temporarily, screenshot each frame beside the source's screenshot at the same size, and fix what differs (spacing, type, a wrong variant, a missing state). Then clear the pin, so the section follows the page's Theme and Mode like everything else.

6. Map and report

When the source is a docs route that shows the page links (website/src/components/PageLinks/) and has no entry yet, add it to pages in website/src/data/figma-nodes.json with the section's node id, then run node scripts/validate-figma-nodes.mjs, so its Figma link lands on the section. A template route is chromeless and shows no page links, so it needs no entry. Report the frames built, the flow and its starting point, any control drawn as a stand-in, and screenshots. Nothing is committed.

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.