figma-agent-brief
/figma-agent-brief
Experimental Banner (ALWAYS display at top of every output)
EXPERIMENTAL SKILL. This skill is a brief-structuring tool for Figma AI (Figma Make, Figma MCP, and sidebar AI actions). Figma AI agent surface area was last validated on 2026-04-11. Figma is shipping changes quarterly. Verify current Figma AI behavior before relying on any output from this skill. If more than 90 days have passed since
experimental_last_validated, treat the output as stale and re-run the Week-1 probe before use.
Purpose
This skill structures a brief that increases the probability Figma AI (primarily Figma Make, secondarily sidebar AI actions) produces something useful on the first attempt. It does not generate design. It does not replace a designer. It does not invoke Figma AI programmatically. It produces a human-readable brief that a PM hands to a designer (or pastes into a Figma Make session) to get a higher-quality first scaffold.
The honest framing: Figma AI in Q1 2026 is a library-awareness-dependent, card-granularity-reliable, brand-voice-unaware tool. This skill makes the brief match what Figma AI can consume and explicitly fences off what it cannot.
When to use
- Briefing Figma Make or Figma's sidebar AI for a card-, section-, or single-screen-sized surface
- The consuming team has a mature, linked design system library (see precondition below)
- You want the first Figma AI scaffold to use the right library components, the right variants, and the right states without hand-correction
- You need to hand the brief to a human designer (the normal case) and want the scope boundaries explicit