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 krusemediallc/uni1-image-ad --skill image-ad-clonegit clone --depth 1 https://github.com/krusemediallc/uni1-image-adWrote 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/krusemediallc/uni1-image-ad/image-ad-clone)<a href="https://agentmods.dev/skills/krusemediallc/uni1-image-ad/image-ad-clone"><img src="https://agentmods.dev/badge/skills/krusemediallc/uni1-image-ad/image-ad-clone/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/krusemediallc/uni1-image-ad/image-ad-clone"><img src="https://agentmods.dev/badge/skills/krusemediallc/uni1-image-ad/image-ad-clone.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 | $0.00126 | $0.03109 |
| Opus 5 | $0.00063 | $0.01554 |
| Sonnet 5 | $0.00025 | $0.00622 |
| Haiku 4.5 | $0.00013 | $0.00311 |
Grade A, and why
image-ad-clone 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 12d 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 — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
image-ad-clone
Take an existing image ad and turn it into a reusable, parameterizable uni-1 prompt template that can be plugged into any brand. Output: a new entry appended to the user's prompt library, ready to be used by uni1-image-ad (or anywhere else) on future generations.
This is the template-creation skill. The companion skill uni1-image-ad is the template-using skill (it generates and uploads ads from filled-in templates). They're meant to be installed together.
Hard rules — never relax
-
Strip platform/screenshot chrome from analysis. When you describe what's in the reference, you're describing the actual ad creative, not the screenshot wrapper. Do not include iOS status bars, "Sponsored"/"Saved" badges, post text/captions surrounding the image, link-card footers, engagement rows, platform tab bars. If the reference is a screenshot of an ad-in-feed, mentally crop the wrapper. The output template must produce a standalone image that would be uploaded as a Meta creative.
-
Always validate by generating. A template that hasn't been round-tripped through uni-1 against the original isn't validated. Run at least one generation with
--image-ref <original>and compare. Refine the prompt until the structure matches. -
Always test the generalized version. Before saving, fill the placeholders with a different brand (use AG1 if the user has no preference — there's a known reference at
iterations/ag1-v2/T1-ios-notes/or the user can supply one) and generate. If the structure breaks, the placeholder set is wrong — fix it. -
Never write brand-specific text into the final template. Wordmarks, product names, slogans, specific photographs, hex colors specific to the source brand — all become
{placeholders}. Only structural content (layout descriptions, photography style, typography family, composition rules) remains literal. -
Save to the user's library, do not silently overwrite. Default save target is
~/.claude/skills/uni1-image-ad/references/prompt-library.mdif it exists; otherwise propose a project-localprompt-library.mdand ask. If the target template name (e.g.,T8 — Apple Notes listicle) collides with an existing entry, ask the user before overwriting.
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.
- 12d ago First seen · 188 lines · 126 tokens per session scan A cf0dd150a3cc
image-ad-clone is a skill published in the GitHub repository krusemediallc/uni1-image-ad (11 stars, last pushed 4mo ago), licensed MIT. It adds 126 tokens to every session and 3,109 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 2026-08-30.
Other skills, from other repositories
claymation-ad-arcads-mcp
Produce an Aardman/Laika-style stop-motion claymation video ad end-to-end through the optional Arcads MCP server: gpt-image-2 storyboards, Seedance 2.0 animation, Arcads TTS, stitching, and subtitles. Use only when the user explicitly requests Arcads MCP or has selected MCP transport and the tools are available in the…
clone-video-ad-seedance-arcads-mcp
Clone an existing video ad onto a new product using Seedance 2.0 through the optional Arcads MCP server. Analyzes the source with arcadsanalyzemedia, adapts its structure, generates via arcadsgeneratevideoseedance20, and polls with arcadswatchasset. Use only when the user explicitly requests Arcads MCP or has selected…
ad-agent-orchestrator
Orchestrates human-approved, end-to-end Meta ad campaigns from competitor research through test concepts, Arcads image/video generation or cloning, human ad copy, PAUSED deployment, explicit activation, and threshold-based performance review. Applies when a user asks to build, launch, test, or iterate a complete…
arcads-mcp
Create and retrieve AI images, videos, and audio through the optional Arcads MCP server (no .env, no curl, no presigned-upload scripts) — Nano Banana / gpt-image / Seedream / Grok images, Seedance 2.0 / Sora 2 / Veo 3.1 / Kling / Grok videos, TTS, talking-head actors, and media editing (stitch, trim, subtitles…
pixar-style-ad-arcads-mcp
Produce a Disney-Pixar-style 3D animated video ad end-to-end through the optional Arcads MCP server: gpt-image-2 storyboards, Seedance 2.0 animation, Arcads TTS, stitching, and subtitles. Use only when the user explicitly requests Arcads MCP or has selected MCP transport and the tools are available in the current…
meta-ad-builder
Research, prepare, and deploy finished creatives to Meta (Facebook/Instagram) through the Marketing API. Lists destinations, optionally creates PAUSED campaign structure, uploads images or videos, creates multi-copy TEXTLIQUIDITY ads in PAUSED state, records lineage, and hands deployed IDs to the read-only performance…