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 scroobius-pip/fudge-mcp --skill fudge-design-researchgit clone --depth 1 https://github.com/scroobius-pip/fudge-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/scroobius-pip/fudge-mcp/fudge-design-research)<a href="https://agentmods.dev/skills/scroobius-pip/fudge-mcp/fudge-design-research"><img src="https://agentmods.dev/badge/skills/scroobius-pip/fudge-mcp/fudge-design-research/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/scroobius-pip/fudge-mcp/fudge-design-research"><img src="https://agentmods.dev/badge/skills/scroobius-pip/fudge-mcp/fudge-design-research.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 | $0.00052 | $0.00632 |
| Opus 5.5 | $0.00021 | $0.00253 |
| Sonnet 5.5 | $0.00010 | $0.00126 |
| Haiku 4.5 | $0.00005 | $0.00063 |
Grade A, and why
fudge-design-research 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 1mo 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 — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Fudge Design Research
Use Fudge as a searchable library of captured websites and their design details.
Start with the live guide
When the connected Fudge MCP server exposes fudge://agent/guide.md, read it before the first Fudge query in a task. The guide describes the current query contract and available evidence. Do not guess a relation, field, or taxonomy label that the guide does not establish.
Choose the right workflow
Find references
For inspiration, critique, or an open-ended design direction:
- Search two or more meaningfully different angles from the request, such as page purpose, layout, component, typography, or visual treatment.
- Pool and deduplicate repeated pages and domains.
- Inspect the strongest available screenshots before describing visible traits.
- Present a small set of genuinely different references and explain what each contributes.
Do not treat one result list as a complete answer when the request is exploratory.
Inspect a website
For a named website or saved page, retrieve evidence for the exact page or domain. Answer with the captured details that matter to the request, such as typography, color roles, spacing, layout, components, borders, shadows, gradients, media, or page context.
Make it clear when evidence comes from a saved capture rather than the current live website. Do not fill missing details with assumptions.
Compare or recommend
Keep hard facts separate from recommendations. Use captured evidence to explain why references differ, then make a recommendation tied to the user's stated goal. Similarity is useful for ranking candidates, not for proving that two designs or fonts are identical.
Review a design
When the user asks for design feedback, search for relevant references before giving substantial advice unless they explicitly want an attachment-only review. Connect each recommendation to a visible problem and a useful example.
Present the result
- Show visual results when the MCP response provides screenshots, pin cards, font previews, palettes, or other rendered evidence.
- Keep the explanation close to the evidence it describes.
- Prefer direct language such as "Fudge found", "the capture uses", and "this reference shows".
- State missing, stale, or unavailable evidence plainly.
- Do not expose internal schema names or retrieval mechanics in ordinary user-facing answers.
- Do not infer font licensing from captured usage or visual similarity.
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.
- 1mo ago First seen · 56 lines · 52 tokens per session scan A ca15615b0549
fudge-design-research is a skill published in the GitHub repository scroobius-pip/fudge-mcp (0 stars, last pushed 1mo ago), licensed MIT. It adds 52 tokens to every session and 632 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-04.
Other skills, from other repositories
flint-interaction-author
Use when: a Flint chart should respond to the reader (highlight, zoom, brush, legend toggle, reorder, annotate, explore, link charts); an application or an agent changes a mounted chart from outside or reads what it shows; or a gesture no preset gives must be built. Produce one valid interactionspec, one ChartUpdate…
flint-theme-author
Use when: creating, translating, refining, reviewing, or validating a custom Flint ThemeSpec from brand guidelines, websites, slide decks, publication references, design tokens, or an existing visual identity. Produce a reusable ThemeSpec JSON object for Flint Theme Lab without inventing fields or changing chart…
b123d-modeling
Use this skill when asked to model, build, or modify a 3D part or assembly with build123d — from a text description, a technical drawing (image or PDF), dimensions in a spec, or an existing STEP/STL file.
Aesthetic Web Design
Apply professional, premium UI/UX design standards when building or styling web interfaces. Use this skill when the user asks to create a website, landing page, web app, dashboard, UI component, or any visual interface — or when they ask to make something look better, more professional, or more beautiful. Covers…
brand-design
Generate brand guidelines, design systems with design tokens (W3C format), creative direction (SCAMPER, Design Thinking, TRIZ), and design briefs with scope and success criteria.
taste-soft
High-end visual design like a premium agency. Defines exact fonts, spacing, shadows, card structures, and animations that make a website feel expensive. Blocks common defaults that make AI designs look cheap or generic. Based on taste-skill by Leon Lin & blueemi, improved by shokunin.