Can I generate images without a campaign brief?

Yes, the skill features a Standalone Mode that allows you to provide a manual prompt, target platform, and aspect ratio via CLI commands.

How does the skill ensure brand consistency?

It reads from a brand-profile.json file to inject specific brand colors, mood, and style presets into the image generation prompts.

Does it provide a quality check for generated images?

Yes, it includes a Quality Gate step where Claude Vision assesses each image against the brief for brand alignment and composition, automatically regenerating assets that score poorly.

What advertising platforms are supported?

The skill includes specific references and templates for Meta (Feed/Reels), TikTok, Google Performance Max, LinkedIn, YouTube, Microsoft, and Apple Ads.

What are the technical requirements to run this skill?

It requires banana-claude (v1.4.1+) with nanobanana-mcp configured. You can set this up by running /banana setup within your terminal.

AI Ad Creative Generator

Name: AI Ad Creative Generator
Author: AgriciDaniel

byAgriciDaniel

•

2,554

•

マーケティングオートメーション

Generates platform-optimized paid advertising images from campaign briefs and brand profiles using automated AI workflows.

This skill automates the production of production-ready ad creatives for major platforms like Meta, Google, TikTok, and LinkedIn. By parsing campaign-brief.md and brand-profile.json, it ensures perfect brand alignment and platform spec compliance across multiple aspect ratios. The tool utilizes a multi-agent architecture to design, validate, and quality-check visuals using Claude Vision, allowing marketing teams and developers to move from strategy to visual assets entirely within the Claude Code environment.

主な機能

012,554 GitHub stars

02Integrated cost estimation and transparency for image generation providers

03Built-in Quality Gate that uses Claude Vision to score and regenerate low-quality assets

04Automated brand injection using brand-profile.json for consistent visual identity

05Parallel agent orchestration for visual design and format validation

06Multi-platform support for Meta, TikTok, Google PMax, LinkedIn, and YouTube specs

ユースケース

01Rapidly prototyping ad creative variations for A/B testing without manual design

02Creating platform-compliant social media thumbnails and banners from text prompts

03Generating a full suite of cross-channel ad assets from a single campaign brief

主な機能

012,554 GitHub stars

02Integrated cost estimation and transparency for image generation providers

03Built-in Quality Gate that uses Claude Vision to score and regenerate low-quality assets

04Automated brand injection using brand-profile.json for consistent visual identity

05Parallel agent orchestration for visual design and format validation

06Multi-platform support for Meta, TikTok, Google PMax, LinkedIn, and YouTube specs

ユースケース

01Rapidly prototyping ad creative variations for A/B testing without manual design

02Creating platform-compliant social media thumbnails and banners from text prompts

03Generating a full suite of cross-channel ad assets from a single campaign brief