Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add marvin-socialista/claude-design-mcp --skill claude-designgit clone --depth 1 https://github.com/marvin-socialista/claude-design-mcpWrote 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/marvin-socialista/claude-design-mcp/claude-design)<a href="https://agentmods.dev/skills/marvin-socialista/claude-design-mcp/claude-design"><img src="https://agentmods.dev/badge/skills/marvin-socialista/claude-design-mcp/claude-design/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/marvin-socialista/claude-design-mcp/claude-design"><img src="https://agentmods.dev/badge/skills/marvin-socialista/claude-design-mcp/claude-design.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high YARA Match · line 150 YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).Fix: Remove the malware payload or compromised file entirely. Investigate how it entered the skill and audit all other artifacts for additional indicators of compromise.
- medium Excessive Agency · line 137 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 152 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00131 | $0.01720 |
| Opus 5 | $0.00066 | $0.00860 |
| Sonnet 5 | $0.00026 | $0.00344 |
| Haiku 4.5 | $0.00013 | $0.00172 |
Grade A, and why
claude-design scanned grade A with 0 findings 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 11d ago.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 161 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Design bridge
Claude Design (claude.ai/design) is a separate product with its own projects, chats and files. This plugin drives it: start projects, send work, see the result, read the code, and take the output.
It runs a real browser on the user's own logged-in session. Every prompt spends
their Claude credits and writes into a real project, so treat send_prompt and
create_project as outward-facing actions, not scratch calls.
Orient first
projectId is required by almost everything and is never guessable.
list_projectsto find the project.kind: "design_systems"filters to design systems.list_chatsto see the threads and which one is active.
send_prompt posts into the active chat. A project usually has more than one
(an app-design thread, a social thread), and the wrong one gives Claude Design
the wrong context. Pass chatId explicitly whenever there is more than one, or
call switch_chat first.
Starting something new
list_templatesshows the tiles on the home screen (Blank, Mobile app design, Slides, Wireframe, Diagram and the rest).create_projectis the home screen's "What should we create?" box: a prompt, an optional template, and it starts building. It returns the newprojectId. Claude Design names the project from the prompt.create_design_systemmakes an empty design-system project. A project's type is fixed at creation, so a design system must be created as one; you cannot convert a project later.link_design_systemsattaches design systems to a project. It replaces the set, so pass everything you want to keep.
Giving it your real code and context
link_local_codeattaches a local folder as the codebase. Point it at the frontend or design-system folder, not a monorepo: it sends text files only, bounded by count and size. The underlying handle is synthetic and lives in the browser page, so re-attach in a later session rather than assuming it persisted.choose_repositorypoints the project at a connected GitHub repo. If GitHub is not connected it will say so; do not try to complete an OAuth grant.upload_figpushes a.figfile from disk.- To reference another project, pass
referenceProjectIdstosend_prompt. There is no menu action for this: the app reads project URLs out of the prompt text, which is what that parameter appends.
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.
- 11d ago First seen · 161 lines · 131 tokens per session scan A 0f6aa715dc2a
claude-design is a skill published in the GitHub repository marvin-socialista/claude-design-mcp (44 stars, last pushed 2d ago), licensed MIT. It adds 131 tokens to every session and 1,720 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
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 $ultragoal 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.