ワンクリックで
brandbook
Generate complete brand identity system for app launch. Strategy to store assets to marketing kit.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Generate complete brand identity system for app launch. Strategy to store assets to marketing kit.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Upgrade DLD framework files from latest template on GitHub.
Autonomous spec execution with subagents (planner, coder, tester, reviewer).
Autonomous spec execution with subagents (planner, coder, tester, reviewer).
Feature specification and research agent. Multi-agent with 4 scouts. Creates specs in ai/features/.
Feature specification and research agent. Multi-agent with 4 scouts. Creates specs in ai/features/.
Manual QA tester — tests product behavior like a real user, not code. Triggers on keywords: test, QA, check behavior, verify feature, manual testing, протестируй, потыкай, проверь как работает
| name | brandbook |
| description | Generate complete brand identity system for app launch. Strategy to store assets to marketing kit. |
| model | opus |
Creates complete brand identity for mobile/web apps through structured dialogue, AI research, and optional MCP image generation.
Activation: brandbook, brandbook {app-name}
Don't use: Redesigning mature brand with existing guidelines (hire an agency)
All files created in brandbook/:
| File | Content | Phases |
|---|---|---|
brand-dna.md | Emotive narrative, convergence shield, archetype, personality, voice | 0-3 |
visual-system.md | Colors (HEX/OKLCH), typography, accessibility validation | 4-6 |
logo-and-icon.md | Concepts, AI prompts for 4 tools, tech specs | 7 |
motion-system.md | Animation guidelines, video specs, motion tokens | 8 |
store-assets.md | Screenshot strategy + copy, Feature Graphic | 9 |
marketing-kit.md | Social specs for 7 platforms, ad strategy, tools | 10 |
prompts.md | All AI prompts collected — copy-paste ready | 11 |
checklist.md | "Zero to Launch" 3-day action plan | 11 |
brand-tokens.css | Tailwind v4 @theme tokens — OKLCH, semantic, dark mode | 12 |
assets/ | Generated images (if MCP available) | 7-8 |
Coder handoff (written outside brandbook/):
| File | Content | Phase |
|---|---|---|
.claude/rules/brand.md | Machine-readable brand rules for coding agents | 13 |
SHIELD: 0. Convergence Shield (emotive narrative, banned defaults, 3 directions)
See: foundation.md
FOUNDATION: 1. Discovery → 2. Brand DNA → 3. Voice Framework
See: foundation.md
DESIGN: 4. Research → 5. Visual System → 6. Accessibility
→ 7. Logo & Icon → 8. Motion
See: design.md
LAUNCH: 9. Store Assets → 10. Marketing Kit
See: launch.md
COMPILE: 11. Compilation + Checklist + MCP Generation
12. Design Tokens Export (brand-tokens.css)
13. Coder Handoff (.claude/rules/brand.md)
See: completion.md
| Module | When to Read | Content |
|---|---|---|
foundation.md | Start | Discovery dialogue, Brand DNA, Voice Framework |
design.md | After foundation | Research, colors, accessibility, logo, motion |
launch.md | After design | Store assets, marketing kit |
completion.md | After all phases | Compilation, checklists, MCP generation, output |
Flow:
SKILL.md → foundation.md → design.md → launch.md → completion.md
biz/, ai/ — product briefs, personas, researchbiz/brandbook/ exists (v1 output). If yes:
mv biz/brandbook/* brandbook/ && rmdir biz/brandbook.claude/rules/brand.md exists (previous run). If yes:
mkdir -p brandbook/assets- MCP auto-saves all generated files to brandbook/assets/ (SAVE_MEDIA_DIR)
- After each generation: show the saved file to user (Read tool)
- Never give CDN URLs — only local paths
- Report: "Saved: brandbook/assets/{filename}"
Step 1: ToolSearch(query: "fal-ai")
Step 2: Tools found? → IMAGE_GEN_MODE = true
Step 3: Not found? → Ask user (AskUserQuestion), do NOT silently fall back
If not found — ask, don't assume:
claude mcp add fal-ai -s user -- uvx --from fal-mcp-server fal-mcpAvailable fal-ai tools (when MCP connected):
IMAGE: generate_image, generate_image_structured, generate_image_from_image,
edit_image, remove_background, upscale_image, inpaint_image,
resize_image, compose_images
VIDEO: generate_video, generate_video_from_image, generate_video_from_video
AUDIO: generate_music
UTIL: list_models, recommend_model, get_pricing, get_usage, upload_file
Use recommend_model tool for AI-powered suggestions, or these defaults:
| Asset | Model | Why |
|---|---|---|
| Logo, icon, text-on-image | fal-ai/recraft-v3 | #1 on leaderboard, vector+raster, best text |
| Photo-realistic imagery | fal-ai/flux-pro/v1.1-ultra | Top photorealism, latest Flux Pro |
| Quick iterations | fal-ai/flux/schnell | $0.003/image, fast |
| Banners with text | fal-ai/recraft-v3 | Superior text rendering |
| Logo animation | generate_video_from_image | Animate logo directly via MCP |
| Background removal | remove_background | Transparent PNG via MCP |
.claude/rules/brand.md)biz/ has product briefs, pre-fill Discovery answersbrandbook/assets/. No CDN links — only local paths. After generation → show image to user