360 Secure SkillHub
360 安全技能中心

AI Brand Analyzer

通过 Gemini Flash 结合 Google 搜索检索品牌官方与广告活动资料,归纳语气、视觉、摄影与传播规范,输出标准品牌身份 JSON,可存文件或自动写入 Ad-Ready 品牌目录,供广告与创意工作流复用。

pauldelavallaz
pauldelavallaz
热度 97.93
最近更新 2026-05-17 11:26:02
分类:营销依赖:代码执行 / 读写文件 / 第三方 API / 内部调用大模型应用场景:市场分析 / 品牌素材 / 研究

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add ai-brand-analyzer
Metadata
name
brand-analyzer
description
Analyze brands to generate comprehensive brand identity profiles (JSON). Use when the user wants to analyze a brand, create a brand profile, or needs brand data for ad generation. Stores profiles for reuse across Ad-Ready, Morpheus, and other creative workflows. Can list existing profiles and update them.

Brand Analyzer: AI Brand Identity Profiler

Analyze any brand to generate a comprehensive brand identity JSON profile using Gemini Flash with Google Search grounding.

Overview

Brand Analyzer creates structured brand identity profiles by:

  1. Researching the brand via Google Search (official data, campaigns, visual identity)
  2. Analyzing brand behavior, visual patterns, photography style, tone of voice
  3. Generating a complete JSON profile following the standard template
  4. Storing the profile for reuse across all creative workflows

When to Use

  • User asks to "analyze a brand" or "create a brand profile"
  • Before running Ad-Ready when the brand isn't in the catalog
  • When the user mentions a brand that doesn't have a profile yet
  • To update/refresh an existing brand profile

Quick Commands

Analyze a brand and save to file

GEMINI_API_KEY="$KEY" uv run {baseDir}/scripts/analyze.py \
  --brand "Brand Name" \
  --output ./brands/Brand_Name.json

Analyze and auto-save to Ad-Ready brands catalog

GEMINI_API_KEY="$KEY" uv run {baseDir}/scripts/analyze.py \
  --brand "Heredero Gin" \
  --auto-save

The --auto-save flag automatically saves to ~/clawd/ad-ready/configs/Brands/{Brand_Name}.json

Print to stdout

GEMINI_API_KEY="$KEY" uv run {baseDir}/scripts/analyze.py --brand "Nike"

Inputs

Input Required Description
--brand ✅ Brand name to analyze
--output Optional Output file path (default: stdout)
--auto-save Optional Auto-save to Ad-Ready brands catalog
--api-key Optional Gemini API key (or set GEMINI_API_KEY env var)

Output Format

The generated JSON follows the standard brand identity template used by Ad-Ready:

{
  "brand_info": { "name", "tagline", "category", "positioning", "vision", "mission", "origin_story" },
  "brand_values": { "core_values", "brand_promise", "differentiators", "non_negotiables" },
  "target_audience": { "demographics", "psychographics" },
  "tone_of_voice": { "personality_traits", "communication_style", "language_register", ... },
  "visual_identity": { "logo", "color_system", "typography", "layout_principles" },
  "photography": { "style", "technical" },
  "campaign_guidelines": { "visual_tone", "model_casting", "product_presentation", ... },
  "brand_behavior": { "do_dont", "immutability" },
  "channel_expression": { "retail", "digital", "print" },
  "compliance": { ... }
}

Integration with Other Workflows

Ad-Ready

Brand profiles are automatically available as brand_profile options when generating ads.

Morpheus Fashion Design

Brand visual identity (colors, photography style, tone) can inform Morpheus campaigns.

Custom Workflows

Load any brand profile JSON to extract visual identity, tone of voice, or campaign guidelines for any creative task.

Analysis Methodology

The analyzer follows a 3-phase approach:

Phase 1: Official Research (via Google Search)

  • Brand website, corporate pages, official communications
  • Locks canonical data: name, founding, positioning, vision, mission, tagline

Phase 2: Campaign Research (via Google Search)

  • Google Images and Pinterest for advertising campaigns
  • Identifies 10+ distinct campaigns
  • Treats them as analytical reference material

Phase 3: Deductive Visual Analysis

  • Cross-sectional analysis of visual patterns
  • Identifies recurring photography style, color systems, typography
  • Fills visual identity fields not covered by official data

API Key

Uses Gemini API. Set via:

  • GEMINI_API_KEY environment variable
  • --api-key flag