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 zstvns/design-skills --skill collateral-designgit clone --depth 1 https://github.com/zstvns/design-skillsWrote 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/zstvns/design-skills/collateral-design)<a href="https://agentmods.dev/skills/zstvns/design-skills/collateral-design"><img src="https://agentmods.dev/badge/skills/zstvns/design-skills/collateral-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/zstvns/design-skills/collateral-design"><img src="https://agentmods.dev/badge/skills/zstvns/design-skills/collateral-design.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 | $0.00184 | $0.10489 |
| Opus 5 | $0.00092 | $0.05244 |
| Sonnet 5 | $0.00037 | $0.02098 |
| Haiku 4.5 | $0.00018 | $0.01049 |
Grade A, and why
collateral-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 10d 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 — 245 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Collateral Design
You apply a brand's defined visual system to the collateral family: decks, one-pagers, print pieces, and ads. This is an Application skill — the color roles, type rules, motif, and mark from creative-direction and logo-design become finished, fixed-format artifacts. It sits beside web-design: same identity, different mediums.
Read design-principles first. Everything here is that skill meeting a page that can't scroll, can't reflow, and often can't be revised after it ships. Hierarchy is the job; restraint protects the message; coherence is not repetition — a deck, a sell sheet, and a feed ad can each look different and still be unmistakably one brand.
Position in the sequence: after
creative-direction(the system),logo-design(the mark), and alongsideweb-design— an Application skill. It consumes the identity; it does not define one. The medium is not one surface but a family of fixed-format deliverables, each with hard constraints web doesn't have: a set trim size, a projector's contrast floor, a thumbnail's legibility, a fold that reorders reading — and no second chance. The piece lands in one shot or not at all.
The Thing That Makes Collateral Different
Web reflows, responds, and ships in commits you can revise tomorrow. Collateral does none of that. Three constraints run through every format here and shape every decision:
- Fixed format. A 16:9 slide, a US-Letter sheet, a 1080×1350 ad, a tri-fold's panel — the frame is set before you start, and the content is designed to it, not poured into whatever height it needs.
- One shot. No hover to reveal, no scroll to continue, no "fold" to push detail below. A projected slide has ~3 seconds; a feed ad has ~1. What doesn't land immediately doesn't land.
- Exported, not live. Collateral ships as a static artifact — a PDF, a sized image, a deck — not a surface you revise tomorrow. It has to stand alone: a deck read with no presenter, a one-pager with no scroll, and — when it's print — CMYK on stock with bleed, not RGB in your editor. Most collateral is screen-first; print is one destination, not the default. If it only works in your tool, with you narrating, it isn't finished.
What ships with it
9 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.
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.
- 10d ago First seen · 245 lines · 184 tokens per session scan A 8da3778598df
collateral-design is a skill published in the GitHub repository zstvns/design-skills (10 stars, last pushed 19d ago), licensed MIT. It adds 184 tokens to every session and 10,489 once invoked, about $0.0009 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
new-template
Scaffolds a new visual (.tsx) in an existing BrandArtisan project, aligned with its brand.md guidelines. Use when the user wants to create a new image/cover/banner/OG in a project ("new visual", "add a cover for X", "create the OG for Y"). Always reads brands/ /brand.md and refuses if it is missing.
brand-assets
Generates or regenerates a project's brand assets (logo, favicon and their variants) through the brand toolchain, faithfully. Use when the user wants to "regenerate the logo / the favicon", "add an icon or logo variant", "a new brand variant", or "set up a project's brand toolchain". Works in tools/ / reusing…
x-page
Creates the X / Twitter profile visuals (profile photo and header / banner) at X's official dimensions. Use when the user wants an "X/Twitter profile photo", a "Twitter header", an "X banner" or "the X avatar" for a project. Produces one or two .tsx templates at the sizes X imposes (profile 400x400, header 1500x500)…
new-project
Initializes a new BrandArtisan visuals project by laying down its reference guidelines. Use when the user wants to create/add a new project, start a new brand or a new visual identity ("new project", "add project X", "set up brand Y"). Creates brands/ /brand.md (mandatory before any visual) and the templates/ / folder.
frame-logo-outro
Segmented logo assembly, glow bloom, and tagline reveal for video outros or brand closing frames.
visual-identity-prompt
Writes a comprehensive visual identity prompt for a show, channel, or publication — a master brief that defines the complete visual language and can be used to art-direct any asset in the project's visual system.