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 agentmods add skills/stablebaseline/mcp/sb-decknpx skills add stablebaseline/mcp --skill sb-deckgit clone --depth 1 https://github.com/stablebaseline/mcpWhat 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 | $0.00116 | $0.02508 |
| Opus 5 | $0.00058 | $0.01254 |
| Sonnet 5 | $0.00023 | $0.00502 |
| Haiku 4.5 | $0.00012 | $0.00251 |
Grade A, and why
sb-deck 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 2d 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 — 204 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Stable Baseline: decks and illustrations
What this skill does
Runs the conversational design agent that builds a polished, on-brand slide deck or a single illustration and places it inside a whiteboard, then exports the result. The conversation is the point: a deck is refined turn by turn rather than regenerated from scratch.
IDs are never hardcoded. Resolve projectId and documentId at runtime from listing
calls in this conversation.
A deck always lives inside a whiteboard
designDeckInWhiteboard and designIllustrationInWhiteboard both require documentId, the
id of an existing whiteboard.
If you do not already have a whiteboard id, ask the user which board to use. Do not create
one automatically. Call createWhiteboard first only when the user explicitly asks for a
brand-new board.
Cost and approval, before anything else
| Action | Cost |
|---|---|
| Build a deck or an illustration | 50 credits |
| Edit an existing one | 10 credits |
| A clarifying question from the agent | free |
| Generated images inside a deck | 5 credits each, off by default |
| Advanced deck building | adds 15 credits per depth level |
Get explicit approval before spending. Call the design tool first without confirm to
receive the exact quote plus the workspace balance, show both to the user in plain terms,
and only call again with confirm: true once they agree. The fee is refunded automatically
if a turn fails or produces no change.
Never call a designer speculatively, and never rerun a build because you did not like the result without asking first.
Building a deck
designDeckInWhiteboard requires documentId and message.
- First call sends the brief. Follow-up calls send the change, the new instruction,
or the answer to the agent's question, together with the
sessionIdordeckIdreturned by the first call. Reusing the session is what makes it a conversation rather than a fresh 50-credit build. - If the brief is ambiguous the agent asks one clarifying question instead of guessing.
Relay that question to the user verbatim, then send their answer back through the same
tool with the same
sessionId. kind: 'deck'is the default premium on-brand HTML deck.kind: 'express'builds a native, deterministic branded executive deck straight onto the board: faster, lower fidelity, one-shot, and not conversational. Say which one you are using and why.slideCounttargets a length.brandKitIdselects a brand kit; see sb-brand.titlenames the deck.
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.
- 2d ago First seen · 204 lines · 116 tokens per session scan A dad0638c470f
sb-deck is a skill published in the GitHub repository stablebaseline/mcp (0 stars, last pushed 7d ago), licensed MIT. It adds 116 tokens to every session and 2,508 once invoked, about $0.0006 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
motion-graphics
A short, design-led motion graphic where motion is the message — kinetic typography, stat count-up, chart/data-viz hit, logo sting / brand lockup, lower-third / callout / social overlay, animated map (highlight regions, connect places, zoom to a location), animated tweet / news-article / headline, webpage / UI…
flier
Design a print-ready, single-page flier as one self-contained HTML file: exactly one fixed-size page.
canvas-design
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.
mockup-device-3d
【模板: 设备 3D 展架 (Device 3D Showcase / HTML-in-Canvas)】 【意图】产品发布、App 演示、设计稿展示。把用户提供的 UI 内容真实渲染到 iPhone / MacBook "屏幕"里, 周围用 CSS 3D transform 模拟 GLTF 模型的玻璃 / 高光 / 折射。Inspired by hyperframes vfx-iphone-device。.
frontend-design
Use this as the design-lead layer for VideoStudio COMPOSE work. It shapes project/composition/composition-manifest.json::artdirection and the model-authored HTML/SVG motion graphics in index.html.
deck-replit
Replit Slides 八套主题 (helix/holm/vance/bevel/world/atlas/bluehouse).