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 S3YED/appie-kit --skill ad-creative-variationsgit clone --depth 1 https://github.com/S3YED/appie-kitWrote 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/s3yed/appie-kit/ad-creative-variations)<a href="https://agentmods.dev/skills/s3yed/appie-kit/ad-creative-variations"><img src="https://agentmods.dev/badge/skills/s3yed/appie-kit/ad-creative-variations/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/s3yed/appie-kit/ad-creative-variations"><img src="https://agentmods.dev/badge/skills/s3yed/appie-kit/ad-creative-variations.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 | $0.00081 | $0.01418 |
| Opus 5 | $0.00041 | $0.00709 |
| Sonnet 5 | $0.00016 | $0.00284 |
| Haiku 4.5 | $0.00008 | $0.00142 |
Grade A, and why
ad-creative-variations 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 8d 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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ad Creative Variations System
Turns one proven-winning ad into a batch of Higgsfield-ready creative variations, using a strict, gated pipeline so every variation preserves what actually makes the ad win instead of guessing from scratch.
Full source spec (Seyed's canonical doc, Version 2): references/ad-creative-system-source.md. Read it once for the complete rationale; this file is the operating summary.
When to use
- The user shares a winning ad (image) and asks for variations, new angles, or scaled creative.
- The user asks to generate Higgsfield prompts from a proven ad.
- The user asks to fix a weak ad angle or generate new hooks for an existing winner.
Hard gate — Step 0
Only proceed if the ad meets the winner bar the user has set (minimum spend AND a ROAS/CPA threshold). Ask for these numbers if not given — do not run the system on unproven ads. If the ad doesn't qualify, say so and stop.
The pipeline
Winning ad
|
Prompt 1 (analyze) -> ad-creative-analyze-winner
| recommendation: PROMPT 2 / PROMPT 3 / BOTH
| text angle: STRONG or WEAK
|
WEAK? -> ad-creative-fix-weak-angle -> take the Top Pick
|
optional -> ad-creative-generate-hooks (Prompt 2 only, max 5 of 15 hooks)
|
ad-creative-new-angles (Prompt 2) and/or ad-creative-new-backgrounds (Prompt 3)
|
Generate each resulting prompt in Higgsfield (Nano Banana 2)
|
QC checklist (below) on every image
|
Save with the naming convention (below) and hand off for review/go-live
Always run in this order. Never skip Prompt 1 — every later step depends on its brief.
Step-by-step
- Analyze — use skill
ad-creative-analyze-winneron the winning ad image. Get RECOMMENDATION (PROMPT 2 / PROMPT 3 / BOTH) and TEXT ANGLE SCORE (STRONG/WEAK). Keep the full brief output — every later step needs it verbatim. - Route:
- TEXT ANGLE is WEAK → run
ad-creative-fix-weak-anglefirst. Prompt 3 is not allowed on a weak text angle (never copy a weak text 5 times). Take the Top Pick forward into step 4. - RECOMMENDATION is PROMPT 3 or BOTH, and text angle is STRONG → Prompt 3 is in play.
- RECOMMENDATION is PROMPT 2 or BOTH → Prompt 2 is in play.
- TEXT ANGLE is WEAK → run
- Optional hooks — if the user wants fresh headline hooks (Prompt 2 route only), run
ad-creative-generate-hooks, then pick at most 5 of the 15 to feed into Prompt 2. - Generate variations:
- Prompt 2 in play → run
ad-creative-new-angles(5 new angles, same visual DNA, format from the brief; use the Top Pick or hooks if produced in steps 2-3). - Prompt 3 in play → run
ad-creative-new-backgrounds(5 new backgrounds, text 100% identical to the brief). - Both in play → run both; treat the result as 10 creatives that go live simultaneously with equal budget per creative.
- Prompt 2 in play → run
- Render each resulting Higgsfield prompt via the Higgsfield skill/CLI (Nano Banana 2 model). One image per prompt.
- QC every image against the checklist below before approving. One NO = regenerate or adjust the prompt; never ship an unchecked image.
- Save + hand off using the naming convention below, and tell the user what to review before it goes live.
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.
- 8d ago First seen · 97 lines · 81 tokens per session scan A c0bc5c86a0d0
ad-creative-variations is a skill published in the GitHub repository S3YED/appie-kit (9 stars, last pushed 16d ago), licensed MIT. It adds 81 tokens to every session and 1,418 once invoked, about $0.0004 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 2026-09-03.
Other skills, from other repositories
browser-edge-cases
SOP for debugging browser automation failures on complex websites. Use when browser tools fail on specific sites like LinkedIn, Twitter/X, SPAs, or sites with Shadow DOM.
aws-patterns
Lambda best practices, S3 event patterns, SQS/SNS fanout, and DynamoDB access patterns for serverless AWS architectures.
review
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/PRD asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use…
agents-md-protocol
Create or review an AGENTS.md file so coding agents get stable repo-local instructions: environment setup, testing, style, security boundaries, PR policy, and handoff rules. Use when a repo lacks durable agent guidance or when a custom harness needs a predictable context file.
investment-memo-generator
Investment memo creation combining financial analysis, document generation, and structured templates. Use when creating investment memos, pitch decks, deal summaries, or investment committee materials.
python-memory-safe-scripts
Memory-safe Python script patterns for long-running processes under systemd MemoryMax constraints. Covers allocator purge (mimalloc/glibc malloctrim), HTTP response lifecycle, DataFrame cleanup, thread-local connection reuse, and periodic GC cadence. Battle-tested through 5 OOM optimization cycles on production GPU…