Skip to content

Add figma-design-extract and design-fidelity-verify skills#1934

Open
jeltehomminga wants to merge 2 commits into
github:stagedfrom
jeltehomminga:add-figma-design-skills
Open

Add figma-design-extract and design-fidelity-verify skills#1934
jeltehomminga wants to merge 2 commits into
github:stagedfrom
jeltehomminga:add-figma-design-skills

Conversation

@jeltehomminga
Copy link
Copy Markdown

Pull Request Checklist

  • I have read and followed the CONTRIBUTING.md guidelines.
  • I have read and followed the Guidance for submissions involving paid services.
  • My contribution adds a new instruction, prompt, agent, skill, or workflow file in the correct directory.
  • The file follows the required naming convention.
  • The content is clearly structured and follows the example format.
  • I have tested my instructions, prompt, agent, skill, or workflow with GitHub Copilot.
  • I have run npm start and verified that README.md is up to date.
  • I am targeting the staged branch for this pull request.

Description

Adds two composable, tool-agnostic design-to-code fidelity skills:

  • figma-design-extract — reads exact values (tokens, sizes, structure) out of a Figma node via the Figma MCP into a build-ready design-spec table mapped to your repo's own design tokens, instead of eyeballing a screenshot.
  • design-fidelity-verify — runs a bounded vision + numeric loop that reads the actually-rendered values off the live app (web getComputedStyle / mobile native view props) and walks every spec row to PASS/FAIL + delta. Web/mobile mechanics live in on-demand references/.

They compose: figma-design-extract produces the spec table, design-fidelity-verify consumes it. Either works standalone.

Both follow the Agent Skills spec (skills/<name>/SKILL.md + bundled assets). Canonical source: jeltehomminga/figma-design-skills (MIT).


Type of Contribution

  • New skill file.

Additional Notes

  • npm run skill:validate passes (all skills valid); docs/README.skills.md regenerated via npm run build.
  • The skill bodies are tool-neutral instruction markdown (no executable code) — they call the Figma MCP and a browser/device MCP, both supported by Copilot's MCP integration.

Two composable, tool-agnostic skills for design-to-code fidelity:

- figma-design-extract: read exact values (tokens, sizes, structure)
  out of a Figma node via the Figma MCP into a build-ready design-spec
  table mapped to your repo's own tokens, instead of eyeballing a
  screenshot.
- design-fidelity-verify: run a bounded vision+numeric loop that reads
  the actually-rendered values off the live app (web getComputedStyle /
  mobile native view props) and walks every spec row to PASS/FAIL+delta.
  Mobile/web mechanics live in on-demand references/.

They compose: extract produces the spec table, verify consumes it.
Canonical source: github.com/jeltehomminga/figma-design-skills (MIT).

Passes npm run skill:validate; docs/README.skills.md regenerated via
npm run build.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Copilot AI review requested due to automatic review settings June 7, 2026 04:03
@github-actions github-actions Bot added new-submission PR adds at least one new contribution skills PR touches skills skill-check-error Skill validator reported errors skill-check-warning Skill validator reported warnings labels Jun 7, 2026
@github-actions
Copy link
Copy Markdown
Contributor

github-actions Bot commented Jun 7, 2026

🔍 Skill Validator Results

⚠️ Warnings or advisories found

Scope Checked
Skills 2
Agents 0
Total 2
Severity Count
--- ---:
❌ Errors 0
⚠️ Warnings 2
ℹ️ Advisories 0

Summary

Level Finding
ℹ️ Found 2 skill(s)
ℹ️ [design-fidelity-verify] 📊 design-fidelity-verify: 2,018 BPE tokens [chars/4: 2,165] (detailed ✓), 17 sections, 1 code blocks
ℹ️ [figma-design-extract] 📊 figma-design-extract: 2,588 BPE tokens [chars/4: 2,673] (standard ~), 18 sections, 2 code blocks
ℹ️ [figma-design-extract] ⚠ Skill is 2,588 BPE tokens (chars/4 estimate: 2,673) — approaching "comprehensive" range where gains diminish.
ℹ️ [figma-design-extract] ⚠ No numbered workflow steps — agents follow sequenced procedures more reliably.
ℹ️ ✅ All checks passed (2 skill(s))
Full validator output ```text Found 2 skill(s) [design-fidelity-verify] 📊 design-fidelity-verify: 2,018 BPE tokens [chars/4: 2,165] (detailed ✓), 17 sections, 1 code blocks [figma-design-extract] 📊 figma-design-extract: 2,588 BPE tokens [chars/4: 2,673] (standard ~), 18 sections, 2 code blocks [figma-design-extract] ⚠ Skill is 2,588 BPE tokens (chars/4 estimate: 2,673) — approaching "comprehensive" range where gains diminish. [figma-design-extract] ⚠ No numbered workflow steps — agents follow sequenced procedures more reliably. ✅ All checks passed (2 skill(s)) ```

Copy link
Copy Markdown
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds two new skills to support a Figma → spec extraction workflow and a measured design-fidelity verification loop (web + mobile references), and registers them in the skills index.

Changes:

  • Added figma-design-extract skill documentation for extracting exact Figma values into a design-spec table.
  • Added design-fidelity-verify skill documentation plus platform reference guides for web and mobile numeric verification.
  • Updated docs/README.skills.md to list/install the new skills.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.

Show a summary per file
File Description
skills/figma-design-extract/SKILL.md New skill doc for extracting Figma structured values into a spec table + token mapping guidance
skills/design-fidelity-verify/SKILL.md New skill doc describing the bounded verification loop and spec-driven numeric checks
skills/design-fidelity-verify/references/verify-web.md Web-specific reference for capture + getComputedStyle/getBoundingClientRect measurements
skills/design-fidelity-verify/references/verify-mobile.md Mobile-specific reference for capture + native/React Native inspection strategies
docs/README.skills.md Registers the two new skills in the documentation index

@@ -0,0 +1,79 @@
# Verify backend — Web
@@ -0,0 +1,55 @@
# Verify backend — Mobile (iOS / Android)
Comment thread skills/figma-design-extract/SKILL.md Outdated
Comment on lines +124 to +126
git diff --unified=0 | grep -nE '^\+' \
| grep -iE '#[0-9a-f]{3,8}\b|rgba?\(|[^a-z-][0-9]+px|font-weight:\s*[0-9]{3}' \
&& echo "raw values found — map them to tokens" || echo "clean"
Comment on lines +45 to +46
width: r.width, height: r.height, // rendered box (sizing)
x: r.x, y: r.y, // position (alignment/placement)
- design-fidelity-verify: replace parent-traversal link to the companion
  skill with the absolute canonical URL (skill-validator forbids `../`
  references that leave the skill directory).
- Rename reference titles "Verify backend — Web/Mobile" → "Verify UI —"
  and the SKILL.md "Pick your backend" heading → "Pick your platform";
  "backend" collided with the B0 "backend/API" preflight note.
- figma-design-extract: narrow the no-raw-values grep gate from `^\+` to
  `^\+[^+]` so it no longer matches `+++ b/file` diff headers.
- verify-web reference: document a ±1px / rounding tolerance for
  getBoundingClientRect fractional pixels to avoid noisy FAILs.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
@github-actions github-actions Bot removed the skill-check-error Skill validator reported errors label Jun 7, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

new-submission PR adds at least one new contribution skill-check-warning Skill validator reported warnings skills PR touches skills

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants