Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/MichelKerkmeester/skilled-agent-harness_spec-driven-loopsnpx agentmods add skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figmaWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figma)<a href="https://agentmods.dev/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figma"><img src="https://agentmods.dev/badge/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figma/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figma"><img src="https://agentmods.dev/badge/skills/michelkerkmeester/skilled-agent-harness_spec-driven-loops/mcp-figma.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00035 | $0.07674 |
| Opus 5 | $0.00017 | $0.03837 |
| Sonnet 5 | $0.00007 | $0.01535 |
| Haiku 4.5 | $0.00003 | $0.00767 |
Grade C, and why
mcp-figma scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- Keywords: figma, figma-cli, figma-ds-cli, figma-desktop, design-tokens, variables, components, render-jsx, figma-export, design-system, figma-daemon, figma-mcp, figma-developer-mcp, code-mode --> How it starts
The opening of the file, as written. The whole thing — 377 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Figma (mcp-figma)
Drive Figma Desktop from the terminal through the silships figma-cli (published as figma-ds-cli) so a coding agent can read, author, modify, and export designs, tokens, and components, and optionally pull design context out of Figma through a Figma MCP via Code Mode. The CLI is the primary surface. The MCP is opt-in. Deep operational detail lives in references/figma-cli-reference.md.
Naming trap (read first). The silships tool publishes to npm as
figma-ds-cli(the unambiguous binary). The npm package literally namedfigma-cliis an UNRELATED tool (unic/figma-cli, binfigma), so nevernpm i -g figma-cli. Thefigma-clicommand only exists when installed from the silships repo (main, exposes bothfigma-ds-cliandfigma-cli). This skill usesfigma-ds-clias the canonical command throughout.Version trap. npm publishes only
[email protected], a minimal build with no--safe, nodaemon, noextractand ~12 commands. The full surface this skill documents (safe connect, daemon, extract/import, ~130 commands) is 1.2.0, available only from the silships repo. Install viascripts/install.sh(--source repo, or--source autowhich upgrades when npm is stale), then confirmfigma-ds-cli --versionis>= 1.2.0.
1. WHEN TO USE
Activation Triggers
Use when the user wants to:
- Drive Figma Desktop from the terminal: create/render frames, components, icons, layouts, or images.
- Work a design system in Figma: tokens, variables, collections,
var:namebinding, shadcn primitives. - Inspect or export from Figma:
extractto DESIGN.md, export PNG/SVG/JSX/CSS/Tailwind/Storybook, screenshots, a11y audits. - Import a design system INTO Figma: from Tailwind config, CSS variables,
tokens.json, or Storybook. - Set up or troubleshoot the figma-cli connection or daemon (safe plugin vs yolo patch, daemon health).
- Optionally pull design context FROM Figma into the agent via a Figma MCP (Framelink
figma) through Code Mode.
What ships with it
54 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/env-template.md 2.6 KB
- assets/utcp-figma-manual.md 4.4 KB
- changelog/v1.0.0.0.md 5.6 KB
- changelog/v1.0.1.0.md 2.3 KB
- changelog/v1.1.0.0.md 2.1 KB
- examples/inspect-export-readonly.sh 2.6 KB runs code
- examples/optional-mcp-context.md 3.4 KB
- examples/README.md 6.6 KB
- examples/safe-connect-daemon-health.sh 2.9 KB runs code
- feature-catalog/a11y-and-analysis/a11y-and-analysis.md 2.4 KB
- feature-catalog/connect-and-daemon/connect-and-daemon.md 4.1 KB
- feature-catalog/design-system-extract-and-import/design-system-extract-and-import.md 2.9 KB
- feature-catalog/export/export.md 2.7 KB
- feature-catalog/feature-catalog.md 19 KB
- feature-catalog/inspect/inspect.md 2.6 KB
- feature-catalog/optional-mcp/optional-mcp-context.md 3.0 KB
- feature-catalog/render-and-create/render-and-create.md 4.2 KB
- feature-catalog/tokens-and-variables/tokens-and-variables.md 3.5 KB
- INSTALL-GUIDE.md 23 KB
- manual-testing-playbook/daemon-health/daemon-status-diagnose.md 5.1 KB
- manual-testing-playbook/detection-setup/binary-detection.md 5.8 KB
- manual-testing-playbook/detection-setup/figma-desktop-required.md 5.1 KB
- manual-testing-playbook/detection-setup/safe-connect.md 5.1 KB
- manual-testing-playbook/intra-routing-recall/connect-daemon.md 5.2 KB
- manual-testing-playbook/intra-routing-recall/create-render.md 5.1 KB
- manual-testing-playbook/intra-routing-recall/design-tokens.md 5.2 KB
- manual-testing-playbook/intra-routing-recall/holdout-create.md 5.4 KB
- manual-testing-playbook/intra-routing-recall/holdout-inspect.md 5.4 KB
- manual-testing-playbook/intra-routing-recall/inspect-export.md 5.1 KB
- manual-testing-playbook/intra-routing-recall/mcp-context.md 5.4 KB
- manual-testing-playbook/intra-routing-recall/negative.md 6.0 KB
- manual-testing-playbook/intra-routing-recall/troubleshoot.md 5.1 KB
- manual-testing-playbook/manual-testing-playbook.md 39 KB
- manual-testing-playbook/optional-mcp/framelink-discovery.md 6.1 KB
- manual-testing-playbook/read-only/read-only-export.md 5.6 KB
- manual-testing-playbook/read-only/read-only-inspect.md 4.6 KB
- manual-testing-playbook/safety-gate/destructive-verb-refused.md 5.7 KB
- mcp-servers/figma-cli/README.md 4.5 KB
- mcp-servers/figma-cli/setup.sh 396 B runs code
- mcp-servers/figma-mcp/README.md 3.0 KB
- README.md 18 KB
- references/figma-cli-reference.md 28 KB
- references/mcp-wiring.md 10 KB
- references/tool-surface.md 26 KB
- references/troubleshooting.md 21 KB
- scripts/_common.sh 1.9 KB runs code
- scripts/connect-safe.sh 1.2 KB runs code
- scripts/connect-yolo.sh 1.4 KB runs code
- scripts/daemon.sh 733 B runs code
- scripts/doctor.sh 3.2 KB runs code
- scripts/install.sh 6.9 KB runs code
- scripts/print-utcp-snippets.sh 1.5 KB runs code
- scripts/README.md 3.4 KB
- scripts/unpatch.sh 486 B runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday Changed 9b535401d41a
- 7d ago First seen · 377 lines · 35 tokens per session scan C 7b6d85ca21fc
mcp-figma is a skill published in the GitHub repository MichelKerkmeester/skilled-agent-harness_spec-driven-loops (35 stars, last pushed yesterday), licensed MIT. It adds 35 tokens to every session and 7,674 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
visual-ralph
Visual Ralph orchestration for frontend UI from generated references, static references, or live URL targets, using $ralph with built-in visual verdict and pixel-diff evidence until the implementation matches and leaves a reproducible design system.
accessibility
Consolidated accessibility skill entrypoint for WCAG 2.2, ARIA Authoring Practices, cognitive accessibility, Section 508, EN 301 549, design intent verification, and the Accessibility Planner workflow.
make-resume
A Chinese-language tool for creating editable HTML resumes that can be changed in a browser and printed to PDF. It uses available resume templates when they are installed and otherwise provides a simpler fallback.