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/lftpadilla/agent-dev-kit/image-finalizenpx skills add LFTPadilla/agent-dev-kit --skill image-finalizegit clone --depth 1 https://github.com/LFTPadilla/agent-dev-kitWhat 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 | $0.00058 | $0.00574 |
| Opus 5 | $0.00029 | $0.00287 |
| Sonnet 5 | $0.00012 | $0.00115 |
| Haiku 4.5 | $0.00006 | $0.00057 |
Grade A, and why
image-finalize 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 yesterday.
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.
What it actually says
Image Finalize Skill
Step 1: Briefing (MANDATORY before generating)
Before generating ANY image, ask the user these key questions:
- What is this for? (Instagram post, banner, logo, etc.)
- What content should it include? (text, services, products, specific messaging)
- What's the focus? (e.g., hosting services, self-hosting, specific product)
- Style preference? (illustration, photo-realistic, minimalist, etc.)
- Logo/branding? (use existing logo, placeholder box, no logo, specific placement)
- Any reference images? (upload or describe a style you like)
Keep it brief — 2-3 questions max, prioritizing the most important for the use case.
Step 2: Iterative Drafts (minimax)
Once briefed, use image_generate with model: "minimax/image-01" for fast, cheap iterations.
- Deliver the image immediately.
- Ask for feedback.
- If logo placeholder needed, include a clearly marked "LOGO" area.
Step 3: Final Polish (gemini)
When user signals finalization via keywords like:
- "final", "done", "approve", "polish", "ready", "finalize"
- "this is good", "love it", "perfect", "yes", "great"
Before proceeding, confirm:
"Ready to generate the final polished version with Gemini 3.1 Pro? This will produce higher quality but may take longer."
If confirmed:
- Use
image_generatewith:model: "google/gemini-3-pro-image-preview"- Pass the last minimax image as reference (
imagesparameter) - If user has a logo file, pass it as reference too
- Prompt: "Polish and finalize this image to production quality. Maintain the same composition and subject. Enhance details, lighting, and overall professional finish. Replace any LOGO placeholder with the provided logo."
- Deliver final image.
Notes
- minimax is fast/cheap for drafts; Gemini is slower/higher quality for final.
- If Google provider shows
configured: no, warn user: "Gemini 3.1 Pro not configured. Please set GEMINI_API_KEY." - Always deliver the final version with
MEDIA:tag for inline display.
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.
- yesterday First seen · 50 lines · 58 tokens per session scan A d86f5e9b9766
image-finalize is a skill published in the GitHub repository LFTPadilla/agent-dev-kit (2 stars, last pushed 4d ago), licensed MIT. It adds 58 tokens to every session and 574 once invoked, about $0.0003 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
design-taste-frontend-v1
The original v1 taste-skill, preserved for projects depending on its exact behavior. The current default is design-taste-frontend (v2 experimental), which is a substantial rewrite. Use this v1 install name only if you need exact backward compatibility.
brandkit
Premium brand-kit image generation skill for creating high-end brand-guidelines boards, logo systems, identity decks, and visual-world presentations. Trained for minimalist, cinematic, editorial, dark-tech, luxury, cultural, security, gaming, developer-tool, and consumer-app brand systems. Optimized for intentional…
high-end-visual-design
Teaches the AI to design like a high-end agency. Defines the exact fonts, spacing, shadows, card structures, and animations that make a website feel expensive. Blocks all the common defaults that make AI designs look cheap or generic.
gpt-taste
Elite UX/UI & Advanced GSAP Motion Engineer. Enforces Python-driven true randomization for layout variance, strict AIDA page structure, wide editorial typography (bans 6-line wraps), gapless bento grids, strict GSAP ScrollTriggers (pinning, stacking, scrubbing), inline micro-images, and massive section spacing.
video-frames
Extract a single frame from a local video at the first frame, a timestamp, or a zero-based frame index using FFmpeg.
storytelling-advisor
Shape pitches, brand stories, presentations, and creative writing with narrative frameworks such as the Hero's Journey, Story Spine, and Freytag's Pyramid. Use when a user wants to structure, critique, or strengthen a story.