AI Content Brief, Script & Outline Generator — Research Assistant for Video & Image generation

AI content brief, script, and outline generator for video, image, and audio production. Works as an AI research assistant that provides expert guidance on workflow design, model selection, parameter optimization, and visual consistency. Generate briefs, scripts, and production outlines before using ima-image-ai, ima-video-ai, ima-voice-ai, or ima-all-ai. Essential reading for planning workflows, choosing models, optimizing prompts and parameters, maintaining character/style consistency, multi-shot production, and avoiding common mistakes. Transforms beginner attempts into professional-grade results. This is a knowledge skill — it provides strategic guidance, not API calls. Covers: content brief generation, script writing, outline creation, AI research assistance, and production best practices for all media types.

Dai Shuo@dai-shuo

Install

openclaw skills install @dai-shuo/ima-knowledge-ai

IMA Knowledge AI

Purpose: This skill provides strategic knowledge to help agents make better decisions when using IMA Studio's content creation APIs. It does NOT make API calls directly — instead, it guides you to use ima-voice-ai, ima-image-ai, ima-video-ai more effectively.

When to Use This Skill

Read this skill BEFORE calling any ima-*-ai skill if you need guidance on:

Workflow Design — How to break down complex user requests into atomic tasks
Model Selection — Which model to choose based on task requirements
Parameter Optimization — How to set parameters for quality, cost, or speed

Example scenarios:

User: "帮我做个宣传视频" → Read workflow-design.md first
User: "用最好的模型生成" → Read model-selection.md to pick the right one
User: "生成16:9的图片" → Read parameter-guide.md for aspect ratio support

Knowledge Structure

All knowledge files live under the references/ directory. When reading from other skills, use paths like ima-knowledge-ai/references/workflow-design.md or ~/.openclaw/skills/ima-knowledge-ai/references/<filename>.md.

This skill contains 12 knowledge modules (8 standalone + 4 modular directories):

Core Knowledge (1-8)

1. workflow-design.md

When to read: Complex user requests that need task decomposition

Task decomposition strategies
Dependency identification (e.g., script → voiceover → video)
Multi-step workflow templates
Common creation patterns

2. model-selection.md

When to read: Choosing between multiple models for a task

Model capability matrix (image/video/voice)
Cost vs. quality trade-offs
Use case recommendations (budget/balanced/premium)
Model limitations and workarounds

3. parameter-guide.md

When to read: Optimizing parameters for a specific task

Resolution/aspect ratio guidelines
Quality vs. speed trade-offs
Common mistakes and fixes
Parameter compatibility matrix

4. visual-consistency.md ⭐ NEW

When to read: Any image/video generation task involving series, characters, or scenes

Why AI generation lacks visual consistency by default
Identifying implicit consistency requirements
Reference image workflow (Image-to-Image / Video-to-Video)
Multi-shot coherence strategies
Common mistakes and best practices

5. video-modes.md ⭐⭐ CRITICAL

When to read: ANY video generation task (MANDATORY before calling ima-video-ai)

image_to_video vs reference_image_to_video (DIFFERENT concepts!)
image_to_video = first frame to video (input becomes frame 1)
reference_image_to_video = reference appearance to video (can change scene)
Traditional two-step vs modern one-step workflow
Fallback strategy when primary method fails
Common mistakes (旺财案例)

6. long-video-production.md 🎬 ESSENTIAL FOR LONG VIDEOS

When to read: User requests video longer than 15 seconds (30s ad, 1min short, 3min promo)

Why models are limited to 10-15 seconds
Multi-shot capability (2-4 camera angles in one generation) 🆕
Three-step workflow: Script → Generate shots → Edit/Stitch
Visual asset preparation (characters, scenes, props)
Shot-by-shot generation strategy
Video editing and stitching techniques
Complete case study: 1-minute fantasy short film

7. character-design.md 🎨 CHARACTER DESIGN / IP DEVELOPMENT

When to read: User needs character design, IP development, game/animation assets, turnaround sheets

Character Design industry overview (games, animation, manga, IP)
Reference-driven workflow (Master Reference → Variants)
Turnaround sheets (front/side/back/3-4 views)
Expression library (happy/angry/sad/surprised...)
Outfit variants (casual/armor/costumes)
Props & weapons reference images
Action poses (idle/walk/run/attack...)
Complete case study: RPG game character "Aria"

8. vi-design.md 🏢 VI DESIGN / BRAND IDENTITY

When to read: User needs VI design, brand identity, logo applications, visual guidelines

VI (Visual Identity) system overview
Foundation system (Logo / Color / Typography / Auxiliary graphics)
Application system (Office / Store / Packaging / Advertising / Digital / Uniform)
Reference-driven workflow (Foundation → Applications)
Logo consistency requirements (highest level)
Color palette management (Primary / Secondary / Neutral / Functional)
Complete case study: "Morning Light Coffee" cafe VI (20+ deliverables)

9. best-practices/ ⭐⭐⭐ COMMERCIAL TEMPLATES (On-Demand)

When to read: Commercial advertising or artistic photography tasks

Structure: Index + 4 scenario files (load only what you need)

README.md — Index with keyword matching (2 KB)
jewelry.md — Jewelry & accessories commercial ads (3 KB)
skincare.md — Skincare & cosmetics commercial ads (3 KB)
perfume.md — Perfume & fragrance commercial ads (3 KB)
cinematic-art.md — Cinematic vintage art photography (4 KB)

Usage: Read index first → Load only relevant scenario file

Token savings: 60-85% compared to loading all scenarios

10. color-theory/ 🎨 COLOR THEORY & CULTURAL SENSITIVITY

When to read: Any design task requiring color selection (logos, posters, brands, products)

Structure: Index + 7 modular files (load only what you need)

README.md — Index with quick navigation (5 KB)
color-psychology.md — 11 major colors' psychology & applications (12 KB) ⭐
color-combinations.md — 5 pairing principles (2 KB)
industry-guide.md — 10 industries' color preferences (1 KB)
cultural-differences.md — 5 regions' basic differences (1 KB)
global-regions.md — 4 regions' detailed guide (4 KB)
religious-systems.md — 5 major religions' color symbolism (5 KB)
application-strategy.md — IMA Studio color decision process (2 KB)

Usage: Read index → Load relevant modules based on task (target region/industry/religion)

Token savings: 70-90% compared to loading all 32 KB

11. design-pitfalls/ 🚫 DESIGN MISTAKES TO AVOID

When to read: Quality assurance for generated content (before final delivery)

Structure: Index + 4 scenario files (29 common mistakes by scene)

README.md — Index with 5 core principles (3 KB)
logo-design.md — Logo design pitfalls (10 rules) (7 KB)
poster-banner.md — Poster/Banner pitfalls (8 rules) (5 KB)
product-ecommerce.md — Product/E-commerce pitfalls (6 rules) (3 KB)
web-ui.md — Web/UI pitfalls (5 rules) (3 KB)

Usage: Read index → Load scenario-specific pitfalls

Token savings: 60-80% compared to loading all 21 KB

12. color-trends-2026/ 📅 2026 COLOR TRENDS

When to read: Design tasks for 2026 market (stay current with trends)

Structure: Index + 5 time/region files (load by current month + target region)

README.md — Index with time-based navigation (6 KB)
annual-colors.md — Pantone Cloud Dancer / WGSN Teal / China Horse Red (4 KB)
spring-summer.md — Mar-Aug trends: Cobalt Blue, Violet, Bright Pink (2 KB)
fall-winter.md — Sep-Feb trends: Dark Luxury theme (1 KB)
regional-differences.md — China/US/Southeast Asia differences (2 KB)
industry-applications.md — Tech/Fashion/Food/Beauty/Home (1 KB)

Usage: Read index → Load current season + target region

Token savings: 75-90% compared to loading all 16 KB

Auto-loading strategy: Check current month → Load relevant season file automatically

Usage Pattern

text

User Request
  ↓
[ima-knowledge-ai] Query relevant knowledge
  ↓
Make informed decision
  ↓
[ima-*-ai] Execute API call with optimized parameters
  ↓
Success!

Example flow: