Generative Media Skills - Setup Guide¶
Source: samuraigpt/generative-media-skills (10K+ combined installs) Category: Media Generation / Video Production Quality Tier: 🟡 Beta
Six generative media skills for end-to-end AI media production. The flagship social-media-video skill reads brand identity documents, ICP profiles, and messaging to produce on-brand social videos - fully optimized for Seedance 2.0's instructional prompt grammar. Also includes logo creation, UI design, and cinematic video direction.
Included Skills¶
| Skill | Installs | Purpose |
|---|---|---|
| muapi-social-media-video | 1.8K | Brand-aware social video (storyboard → reference frames → Seedance 2.0) |
| muapi-seedance-2 | 1.7K | Seedance 2.0 video generation with instructional prompt grammar |
| muapi-logo-creator | 1.7K | AI logo generation from brand identity docs |
| muapi-ui-design | 1.7K | UI/screen design generation |
| muapi-nano-banana | 1.6K | Image generation (model-agnostic) |
| muapi-cinema-director | 1.6K | Cinematic video direction with camera movement scripting |
Installation¶
npx skills add samuraigpt/generative-media-skills --skill muapi-social-media-video
npx skills add samuraigpt/generative-media-skills --skill muapi-seedance-2
npx skills add samuraigpt/generative-media-skills --skill muapi-logo-creator
npx skills add samuraigpt/generative-media-skills --skill muapi-ui-design
npx skills add samuraigpt/generative-media-skills --skill muapi-nano-banana
npx skills add samuraigpt/generative-media-skills --skill muapi-cinema-director
Prerequisites¶
| Requirement | Details |
|---|---|
| Brand Files | brand-identity.md, ICP.md, messaging.md in working directory or brand/ subdirectory |
| Seedance 2.0 | Required for video generation skills (muapi-seedance-2, muapi-social-media-video) |
| Image Model | Any available image model for reference frame generation |
Key Capabilities¶
Social Media Video Creator¶
End-to-end pipeline: Brand Files → Storyboard → Reference Images → Seedance 2.0 Video:
- Read brand files: Extracts visual style, color palette, target audience, core messaging
- Write social post + storyboard: Hook line, body, hashtags, CTA + time-coded storyboard
- Generate reference frames: Uses best available image model for key scenes
- Produce video: Optimized Seedance 2.0 prompt with camera movement, transitions
Cinema Director¶
Cinematic video direction with camera movement scripting - dolly, pan, zoom, tracking shots - for professional-quality AI video.
Logo Creator¶
Reads brand identity documents and generates logo variations optimized for different use cases (social profile, website header, app icon).
UI Design¶
Generates UI/screen designs from product descriptions, brand identity, and platform context.
Quick Start - Hermes Agent¶
# Create sample brand files
mkdir -p brand
cat > brand/brand-identity.md << 'EOF'
# Brand: CorpusIQ
## Visual Style
- Colors: #0f0f12 (dark), #f0b90b (gold accent)
- Typography: Modern, technical
- Logo: Geometric owl silhouette
## Tone
Technical, operator-focused, helpful-first
EOF
cat > brand/ICP.md << 'EOF'
# ICP: Business Operators
- Role: Operations managers, founders, growth leads
- Pain: Too many disconnected AI tools, need unified platform
- Goal: 10x operational efficiency with AI
EOF
cat > brand/messaging.md << 'EOF'
# Core Messaging
- Hook: "One platform. Every AI tool. Zero complexity."
- CTA: "Start free at corpusiq.ai"
- Tagline: "Operate at the speed of AI"
EOF
# Install the social media video skill
npx skills add samuraigpt/generative-media-skills --skill muapi-social-media-video
Verification¶
npx skills list | grep samuraigpt
# Should list installed generative media skills
Notes¶
- Brand-file driven: Skills read markdown brand documents - no coding required for consistent output
- Seedance 2.0: Primary video engine - uses instructional prompt grammar for precise camera control
- Use case for CorpusIQ: UGC video production, social media content, logo refinement, UI mockups for product specs
- Model agnostic: Image generation uses whatever model is available - not locked to one provider
- Related skills: corpusiq-media-pipeline, corpusiq-heygen-video-automation, hyperframes
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.