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 hassancs91/claude-image-generation --skill level-3-image-generatorgit clone --depth 1 https://github.com/hassancs91/claude-image-generationWrote 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/hassancs91/claude-image-generation/level-3-image-generator)<a href="https://agentmods.dev/skills/hassancs91/claude-image-generation/level-3-image-generator"><img src="https://agentmods.dev/badge/skills/hassancs91/claude-image-generation/level-3-image-generator/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/hassancs91/claude-image-generation/level-3-image-generator"><img src="https://agentmods.dev/badge/skills/hassancs91/claude-image-generation/level-3-image-generator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00156 | $0.00819 |
| Opus 5.5 | $0.00062 | $0.00328 |
| Sonnet 5.5 | $0.00031 | $0.00164 |
| Haiku 4.5 | $0.00016 | $0.00082 |
Grade A, and why
cf-image 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.
cf-image
Turn a natural-language request into a generated image using Cloudflare Workers AI (@cf/black-forest-labs/flux-1-schnell). This skill is the "connect Claude to image generation" loop: the user describes what they want, Claude does the prompt engineering, the bundled script calls the Cloudflare API, and the image is saved locally.
When this skill applies
The user wants an image generated from a description — e.g. "generate a logo idea for a coffee shop", "make a thumbnail for my AI video", "picture of a cyberpunk cat". Claude is responsible for turning a vague ask into a good prompt before generating.
It does not apply to: editing an existing image, vector/SVG art, or video.
How to run it
From the repo root:
python .claude/skills/level-3-image-generator/generate.py "<optimized prompt>" -o <filename>.jpg
"<optimized prompt>"— the prompt you write (see below), not the user's raw words.-o/--output— where to save (defaultout.jpg). Pick a short descriptive name, e.g.coffee-logo.jpg.--steps— diffusion steps (default4). flux-schnell is tuned for few steps; 4–8 is plenty. Higher is slower, not always better.
The script prints Saved <file> ... on success, or a Cloudflare API error message on failure.
The one job Claude must do: write a strong prompt
The raw script would need a hand-written prompt. The value of this skill is that Claude expands the user's intent into a vivid, specific prompt before calling the script. A good Flux prompt names:
- Subject — what's in the frame, concretely.
- Style/medium — photo, 3D render, flat vector, watercolor, isometric, etc.
- Composition & framing — close-up, wide shot, centered, rule-of-thirds.
- Lighting & mood — golden hour, neon, soft studio light, dramatic shadows.
- Color/detail cues — palette, background, level of detail.
Example transformation:
- User: "a logo for a coffee shop"
- Prompt you pass: "minimalist logo of a steaming coffee cup, flat vector illustration, warm brown and cream palette, centered on a clean off-white background, simple bold shapes, modern branding"
What ships with it
1 file 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.
- 1mo ago First seen · 56 lines · 156 tokens per session scan A 02b5181e8e7d
cf-image is a skill published in the GitHub repository hassancs91/claude-image-generation (109 stars, last pushed 2mo ago), licensed MIT. It adds 156 tokens to every session and 819 once invoked, about $0.0006 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 2026-08-30.
Other skills, from other repositories
app-mockup
Create deterministic browser, tablet, laptop, and phone mockups from real screenshots via the local TypeScript renderer in this plugin. Use for marketing visuals, store screenshots, and framed UI previews without ML image generation.
edge-commerce
Reduce latency globally by running geo-routing, A/B tests, and personalization logic at the network edge using Cloudflare Workers or Vercel.
review-stack
Stack review of one or more third-party tools (OXC, TanStack, Vite, cf, Sentry, pnpm, Claude Code...) against current official best practices, or of what a dependency bump brings (bumped): one parallel subagent per tool, verified findings, one consolidated plan in chat. USE WHEN: user runs /review-stack or explicitly…
cf
USE WHEN: running, finding, or troubleshooting Cloudflare CLI (cf) commands in this repo - dev servers, builds and their Build Output, Env types, startup profiling, production profiles, Workers Issues, and telemetry, local KV/D1/R2 data, the commands that still need Wrangler, or the cf equivalent of a Wrangler habit.…
motion-graphics
Use when the user wants a short, design-led motion graphic where motion is the message: kinetic typography, stat or number count-up, chart/data-viz hit, logo sting, brand lockup, lower-third, callout, social overlay, animated headline/tweet/news item, motion poster, or quick captured-page highlight. Usually under 10s…
website-to-video
Capture a general website/URL and turn it into a HyperFrames video (site tour, showcase, or social clip from the site's own visuals). Uses headless Chrome screenshots + brand assets. Use when intent is general — portfolio/blog/landing-page showcase or social clip from the site. NOT for: product/SaaS launch or promo (→…