Content & Writing  ·  Official

figma-use

**MANDATORY prerequisite** — you MUST invoke this skill BEFORE every `use_figma` tool call.


Composite

3.9

C 4.9 · A 3.2

How we got there

Craft · D1–D5

D1 · Trigger clarity 5.0
D2 · Output specificity 5.0
D3 · Scope precision 5.0
D4 · Self-containment 5.0
D5 · Reusability 4.0

Adoption · A1–A5

A1 · Maintenance 2.5
A2 · Documentation 4.3
A3 · License 2.5
A4 · Adoption 4.3
A5 · Authorship 2.0

02 — Review

Our evaluation


Tier-2 Review: figma-use (Score: 4.8 / 5.0)

What We Attempted

We attempted to validate the figma-use skill, which provides structured rules and reference material for executing JavaScript in Figma files via the use_figma MCP tool. The skill is designed to be loaded before every use_figma call, providing critical API usage rules, reference file pointers, and best practices for manipulating Figma documents programmatically. Our test harness attempted to install the skill and run a smoke-invocation to verify it could be loaded and provide its intended guidance.

What Failed

0 tests passed, 0 partial, 0 failed, 2 skipped. The skill could not be tested in our environment because it is fundamentally a set of rules and reference pointers, not an executable artifact.

  • Install (skipped): No install command is documented in SKILL.md. The skill is a collection of Markdown rules and reference files (.md, .d.ts) — not a Python package, CLI tool, or any installable asset. Our harness found nothing to install.
  • Smoke-invocation (partial): The skill requires the use_figma MCP tool to be available in the runtime environment, along with actual Figma API access. Neither was present in our test harness. The skill itself is purely instructional — it cannot be "invoked" in the traditional sense. It provides guidance for humans or LLMs to write correct use_figma calls, but does not contain executable code.

What We Observed

The skill is well-structured for its intended purpose. It clearly documents 13 critical rules for using the Figma Plugin API, including subtle gotchas like:

  • Colors must be 0–1 range, not 0–255
  • figma.notify() throws "not implemented"
  • layoutSizingHorizontal/Vertical = 'FILL' must be set after parent.appendChild(child)
  • Fonts must be loaded before text operations

It references five supporting files (plugin-api-standalone.index.md, plugin-api-standalone.d.ts, wwds.md, gotchas.md, variable-patterns.md) that would provide additional depth in a real Figma environment.

The skill also correctly identifies its own dependency: it must be used alongside figma-generate-design for full page-building workflows, and it explicitly warns against skipping the skill before use_figma calls.

Rating Disclaimer

The composite score of 4.8 / 5.0 is theoretical and based solely on static analysis of the skill document. It received perfect scores on trigger clarity, output specificity, scope precision, and self-containment because the document is internally consistent, well-organized, and clearly scoped. The 4.0 on reusability reflects its narrow domain (Figma Plugin API only). This rating cannot be confirmed until the skill is physically re-run in an environment with the use_figma MCP tool and Figma API access. The D-048 test harness results document this limitation transparently.

Is This Skill Still Valuable in Principle?

Yes, absolutely. The figma-use skill addresses a genuine pain point: the Figma Plugin API has many non-obvious rules and failure modes (fonts not loaded, colors in wrong range, layout properties set in wrong order). Having a structured, reusable set of rules that must be loaded before every use_figma call is an excellent design pattern for reducing LLM errors in programmatic Figma manipulation. The skill's value is entirely in its instructional and rule-enforcement role — it prevents common bugs before they happen. For any team using LLMs to generate Figma Plugin API code, this skill would be a high-value addition to their workflow. The test failures are entirely an artifact of our harness's limitations, not a reflection on the skill's quality or usefulness.

03 — Tests

What we tried


Tests simulated against README claims; pending physical re-run in Docker harness. Ran 2026-05-21.

Overall: broken. 0 tests passed, 0 partial, 0 failed, 2 skipped; key blocker: skill is a set of rules and references for using the Figma Plugin API via an MCP tool, not a standalone executable.

Inferred dependencies: MCP tool 'use_figma', Figma Plugin API (browser/plugin context), references/plugin-api-standalone.index.md, references/plugin-api-standalone.d.ts, references/working-with-design-systems/wwds.md, references/gotchas.md, references/variable-patterns.md.

Test Status Notes
install skipped No install command documented in SKILL.md; skill is a set of rules and references, not a Python package.
smoke-invocation partial Requires MCP tool 'use_figma' which is not available in this environment; also requires Figma API access. The skill itself is a set of rules, not executable code.
04 — Cross-validation

1 source verified

Install

Use this skill

/plugin install figma-use