{"id":4239,"date":"2026-09-03T17:34:51","date_gmt":"2026-09-03T09:34:51","guid":{"rendered":"https:\/\/ai.jxgzhc.cn\/?p=4239"},"modified":"2026-09-03T17:34:55","modified_gmt":"2026-09-03T09:34:55","slug":"%e4%b8%baai-agent%e6%89%93%e9%80%a0%e7%9a%84%e7%94%9f%e6%88%90%e5%bc%8f%e5%aa%92%e4%bd%93%e6%8a%80%e8%83%bd%e9%9b%86%e5%90%88","status":"publish","type":"post","link":"https:\/\/ai.jxgzhc.cn\/?p=4239","title":{"rendered":"\u4e3aAI Agent\u6253\u9020\u7684\u751f\u6210\u5f0f\u5a92\u4f53\u6280\u80fd\u96c6\u5408"},"content":{"rendered":"

\"\u4e3aAI<\/p>\n

\u5f00\u7bb1\u5373\u7528\u83b7\u53d6\u4e13\u4e1a\u7ea7AI\u751f\u6210\u56fe\u50cf\u3001\u89c6\u9891\u3001\u97f3\u9891\u5185\u5bb9\uff0c\u5feb\u901f\u5b8c\u6210\u8425\u9500\u3001\u521b\u4f5c\u7c7b\u5a92\u4f53\u9879\u76ee\uff0c\u5927\u5e45\u63d0\u5347\u521b\u4f5c\u6548\u7387<\/strong><\/p>\n

\u8be6\u7ec6\u4ecb\u7ecd<\/h2>\n

\ud83c\udfad Generative Media Skills for AI Agents<\/p>\n

The Ultimate Multimodal Toolset for Claude Code, Cursor, Gemini CLI, and OpenCode.
A high-performance, schema-driven architecture for AI agents to generate, edit, and display professional-grade images, videos, and audio \u2014 powered by the muapi-cli .<\/p>\n

\ud83d\ude80 Get Started | \ud83c\udfac Recipe Pack | \ud83c\udfa8 Expert Library | \u2699\ufe0f Core Primitives | \ud83e\udd16 MCP Server | \ud83d\udcd6 Reference<\/p>\n

Related Projects<\/p>\n

MiniMax-H3-API \u2014 Python SDK for MiniMax H3 video-generation workflows on Muapi.<\/p>\n

awesome-minimax-h3-prompts \u2014 Prompt gallery and runnable examples for the MiniMax H3 skills.<\/p>\n

Wan-3.0-API \u2014 Python SDK and MCP server for Wan 3.0-compatible video-generation workflows.<\/p>\n

Open-Generative-AI \u2014 Free self-hosted AI media studio \u2014 GUI alternative to these skills for the same model set<\/p>\n

Awesome-GPT-Image-2-API-Prompts \u2014 Curated GPT-Image-2 prompts to use with these skills<\/p>\n

Awesome-Gemini-Omni-API-Prompts \u2014 Curated Gemini Omni prompts for video generation<\/p>\n

AI-Voice-Agent \u2014 Self-hosted AI voice agent for real-time voice conversations, sales calls, and customer support<\/p>\n

awesome-ai-image-models \u2014 compare AI image models by API, price & quality<\/p>\n

flux-3-video-api \u2014 Python wrapper focused on FLUX 3 Text-to-Video and Image-to-Video<\/p>\n

ai-creator-academy \u2014 free curriculum teaching creators to monetize generative AI, built on these same skills<\/p>\n

Flux-3-Dev-API \u2014 Python wrapper for Black Forest Labs' FLUX 3 (Dev variant) \u2014 text-to-image, image-to-image, text-to-video, image-to-video<\/p>\n

Grok-Imagine-Image-2-API \u2014 Python SDK and MCP server for Grok Imagine Image 2.0 generation and editing through MuAPI<\/p>\n

awesome-flux-3-api-prompts \u2014 FLUX 3 API guide, prompts, and parameters<\/p>\n

seedance-2.5-mcp \u2014 MCP server for generating Seedance 2.5 Preview videos through MuAPI.<\/p>\n

seedance-2-mcp \u2014 MCP server for generating Seedance 2 videos through MuAPI.<\/p>\n

\u2728 Key Features<\/p>\n

\ud83e\udd16 Agent-Native Design \u2014 CLI-powered scripts with structured JSON outputs, semantic exit codes, and –jq filtering for seamless agentic pipelines.<\/p>\n

\ud83e\udde0 Expert Knowledge Layer \u2014 Domain-specific skills that bake in professional cinematography, atomic design, and branding logic.<\/p>\n

\u26a1 CLI-Powered Core \u2014 All primitives delegate to muapi-cli \u2014 no curl, no JSON parsing, no boilerplate.<\/p>\n

\ud83d\uddbc\ufe0f Direct Media Display \u2014 Use the –view flag to automatically download and open generated media in your system viewer.<\/p>\n

\ud83d\udcc1 Local File Support \u2014 Auto-upload images, videos, faces, and audio from your local machine to the CDN for processing.<\/p>\n

\ud83c\udf08 100+ AI Models \u2014 One-click access to Midjourney v7, Flux Kontext, Seedance 2.0, Kling 3.0, Veo3 , and more.<\/p>\n

\ud83d\udd0c MCP Server \u2014 Run muapi mcp serve to expose all 19 tools directly to Claude Desktop, Cursor, or any MCP-compatible agent.<\/p>\n

\ud83c\udfd7\ufe0f Scalable Architecture<\/p>\n

This repository uses a Core\/Library split to ensure efficiency and high-signal discovery for LLMs:<\/p>\n

\u2699\ufe0f Core Primitives ( \/core )<\/p>\n

Thin wrappers around muapi-cli for raw API access.<\/p>\n

core\/media\/ \u2014 File upload<\/p>\n

core\/edit\/ \u2014 Image editing (prompt-based)<\/p>\n

core\/platform\/ \u2014 Setup, auth & result polling<\/p>\n

\ud83d\udcda Expert Library ( \/library )<\/p>\n

High-value skills that translate creative intent into technical directives.<\/p>\n

Cinema Director ( \/library\/motion\/cinema-director\/ ) \u2014 Technical film direction & cinematography.<\/p>\n

Nano-Banana ( \/library\/visual\/nano-banana\/ ) \u2014 Reasoning-driven image generation (Gemini 3 Style).<\/p>\n

UI Designer ( \/library\/visual\/ui-design\/ ) \u2014 High-fidelity mobile\/web mockups (Atomic Design).<\/p>\n

Logo Creator ( \/library\/visual\/logo-creator\/ ) \u2014 Minimalist vector branding (Geometric Primitives).<\/p>\n

Seedance 2 (Doubao Video) ( \/library\/motion\/seedance-2\/ ) \u2014 Director-level cinematic video generation with text-to-video, image-to-video, and video extension with native audio-video sync.<\/p>\n

AI Clipping ( \/library\/edit\/ai-clipping\/ ) \u2014 Long video \u2192 ranked vertical short clips in one managed API call. Server-side transcription, virality ranking, dedupe, and face-tracked auto-crop \u2014 no local Whisper or LLM.<\/p>\n

YouTube Shorts ( \/library\/social\/youtube-shorts\/ ) \u2014 Platform-aware preset over AI Clipping (Shorts \/ TikTok \/ Reels \/ Feed defaults).<\/p>\n

Plus 41 ready-to-run workflow recipes organized by output type \u2014 see \ud83c\udfac Recipe Pack below.<\/p>\n

\ud83c\udfac Recipe Pack<\/p>\n

Forty-one LLM-orchestrated workflow recipes that combine multiple muapi-cli calls into named end-to-end pipelines (e.g. photo of person \u2192 3D action figure , product photo \u2192 cinematic 10s ad ). Each skill is a SKILL.md the agent reads and follows; bring your own consuming agent (Claude Code, Cursor, MCP) \u2014 these are recipes, not bash wrappers.<\/p>\n

Motion \/ Video (16)<\/p>\n

Skill
Description<\/p>\n

3D Logo Animation
Transform a 2D logo into a premium 3D version and animate it with professional cinematic effects<\/p>\n

AI Fight Scene Generator
High-cut-density action \/ fight scene \u2014 16-cell storyboard image drives Seedance 2.0 i2v for shot-by-shot choreography<\/p>\n

Animal Vlogger Video
Hilarious, ultra-realistic anthropomorphic-animal vlogger acting like a human in a real-world setting<\/p>\n

Cartoon Dance Animation
Convert a photo into a Pixar-style 3D cartoon, then animate using a reference dance\/motion video<\/p>\n

Character Story Video
Multi-part animated story video \u2014 establish a consistent character then animate sequential scenes<\/p>\n

Drone-Style Video
Aerial drone-perspective footage \u2014 bird's-eye sweeps, orbit shots, and flyover sequences<\/p>\n

Giant Product Showcase
Dramatic giant-scale product visual (building-sized object next to a person), optionally animated<\/p>\n

Jewelry Product Video
Luxury jewelry ad with high-end commercial cinematography and detailed macro animation<\/p>\n

Music Video
Short music video from a song theme \u2014 keyframes, animation per beat, matching music track<\/p>\n

One-Shot Video
Single continuous cinematic shot \u2014 no cuts, one seamless flowing scene<\/p>\n

Cinematic Product Ad
Cinematic 5\u201310s product ad from a product photo + brand brief<\/p>\n

Product Showcase Video
Dynamic product showcase with explosive ingredient arrangement + realistic motion animation<\/p>\n

Product Video Ad Maker
High-end cinematic product video ad starting from a simple product photo<\/p>\n

Talking Baby Video
Viral-style talking-baby video with custom costumes and scripts<\/p>\n

UGC Lifestyle Try-On
UGC-style lifestyle photos & video of a person using your product \u2014 authentic, social-native<\/p>\n

UGC Video Factory
Person photo + product photo + script \u2192 10s vertical 9:16 UGC video ad with native dialogue (Nano-Banana Pro Edit \u2192 Seedance 2.0 VIP i2v)<\/p>\n

Social (5)<\/p>\n

Skill
Description<\/p>\n

Instagram Post
Polished on-brand Instagram post \u2014 hero image + caption + hashtags<\/p>\n

Product Campaign Pack
Full multi-channel campaign \u2014 hero visuals, social assets, short ad video, platform crops<\/p>\n

RedNote Cover
Xiaohongshu (\u5c0f\u7ea2\u4e66) cover image \u2014 vibrant lifestyle aesthetic with typography overlay<\/p>\n

Social Media Pack
Re-render a hero image into Instagram \/ TikTok \/ Shorts \/ X aspect ratios<\/p>\n

UGC Ads Workflow
UGC video ad pipeline \u2014 combine selfie + product image, write script, animate<\/p>\n

Visual \/ Images & Design (21)<\/p>\n

Skill
Description<\/p>\n

Action Figure Generator
Convert a photo of a person into a custom 3D action figure with collectible toy packaging<\/p>\n

Ad Creative Set
High-converting ad set \u2014 hero image, copy variations, platform crops for Meta \/ Google \/ LinkedIn<\/p>\n

Amazon Product Listing Pack
Full Amazon listing image set \u2014 hero, lifestyle, infographic, comparison\/detail closeups<\/p>\n

Blog Header
Professional 1200\u00d7628 blog header image with optional title composition guidance<\/p>\n

Brand Kit
Cohesive brand visual kit \u2014 logo concept, color palette, typography pairings<\/p>\n

Brochure Designer
Multi-page brochure \u2014 cover, inner spread, back \u2014 for business, real estate, events, launches<\/p>\n

Couple Grid Creator
Stylized 6-box grid of a couple in romantic poses, each pose framed inside cardboard packaging<\/p>\n

Brand Design Guide
Comprehensive design guide \u2014 palette, typography, UI components, visual identity rules<\/p>\n

Fashion Try-On
Virtually try outfits by combining a person's photo + clothing item, optional fashion model video<\/p>\n

Floor Plan Rendering
Design a 2D floor plan and convert into a realistic 3D architectural rendering<\/p>\n

Interior Design
Pro interior design visualizations \u2014 redesign rooms, generate concepts, visualize furniture styles<\/p>\n

Interior Design Visualizer
Generate an empty room and fill it with stylish furniture \/ decor; or redesign an existing room<\/p>\n

Keyboard Art Maker
Artistic top-down photos of keyboard keycaps arranged to spell custom messages<\/p>\n

Logo + Branding Package
Logo + full branding package \u2014 variations (dark\/light\/icon), palette, mockups<\/p>\n

Logo Generator
Quick single-shot polished logo \u2014 fast, clean vector aesthetic with accurate brand-name text<\/p>\n

Multi-Angle Reshoot
Re-render a subject from dramatic camera angles (fish-eye, bird's-eye, low, macro) \u2014 identity preserved<\/p>\n

Multi-Angle Shots
Full multi-angle product shot set \u2014 front, side, back, top-down, 45\u00b0<\/p>\n

Selfie with Celebrities
Realistic behind-the-scenes selfie of the user with a celebrity; optional cinematic long-take<\/p>\n

Storyboard Generator
Generate N keyframes for a short story or scene sequence (image only, no video)<\/p>\n

URL to Design
Analyze a website URL and generate a redesigned, improved UI with modern aesthetics<\/p>\n

YouTube Thumbnail
High-CTR YouTube thumbnail \u2014 striking imagery, bold text placement, emotional face\/subject<\/p>\n

Each recipe declares its inputs and a Steps body. Pass the inputs and let your agent execute the steps via muapi CLI calls (or raw API for endpoints that don't yet have a CLI alias \u2014 see the per-skill Notes for the Executing Agent footer).<\/p>\n

\ud83d\ude80 Quick Start<\/p>\n

1. Install the muapi CLI<\/p>\n

The core scripts require muapi-cli . Install it once:<\/p>\n

# via npm (recommended \u2014 no Python required)
npm install -g muapi-cli<\/p>\n

# via pip
pip install muapi-cli<\/p>\n

# or run without installing
npx muapi-cli –help<\/p>\n

2. Configure Your API Key<\/p>\n

# Interactive setup
muapi auth configure<\/p>\n

# Or pass directly
muapi auth configure –api-key "YOUR_MUAPI_KEY"<\/p>\n

# Get your key at https:\/\/muapi.ai\/dashboard?utm_source=github&utm_medium=readme&utm_campaign=generative-media-skills<\/p>\n

3. Install the Skills<\/p>\n

# Install all skills to your AI agent
npx skills add SamurAIGPT\/Generative-Media-Skills –all<\/p>\n

# Or install a specific skill
npx skills add SamurAIGPT\/Generative-Media-Skills –skill muapi-media-generation<\/p>\n

# Install to specific agents
npx skills add SamurAIGPT\/Generative-Media-Skills –all -a claude-code -a cursor<\/p>\n

4. Generate Your First Image<\/p>\n

muapi image generate "a cyberpunk city at night" –model flux-dev<\/p>\n

# Download the result automatically
muapi image generate "a sunset over mountains" –model hidream-fast –download .\/outputs<\/p>\n

# Extract just the URL (agent-friendly)
muapi image generate "product on white bg" –model flux-schnell –output-json –jq '.outputs[0]'<\/p>\n

5. Run an Expert Skill<\/p>\n

# Use Nano-Banana reasoning to generate a 2K masterpiece
bash library\/visual\/nano-banana\/scripts\/generate-nano-art.sh \\
–file .\/my-source-image.jpg \\
–subject "a glass hummingbird" \\
–style "macro photography" \\
–resolution "2k" \\
–view<\/p>\n

6. Direct a Cinematic Scene<\/p>\n

cd library\/motion\/cinema-director<\/p>\n

# Create a 10-second epic reveal
bash scripts\/generate-film.sh \\
–subject "a cybernetic dragon over Tokyo" \\
–intent "epic" \\
–model "kling-v3.0-pro" \\
–duration 10 \\
–view<\/p>\n

# Animate a reference image into video
bash library\/motion\/seedance-2\/scripts\/generate-seedance.sh \\
–mode i2v \\
–file .\/concept.jpg \\
–subject "camera slowly pulls back to reveal the full landscape" \\
–intent "reveal" \\
–view<\/p>\n

# Extend an existing video
bash library\/motion\/seedance-2\/scripts\/generate-seedance.sh \\
–mode extend \\
–request-id "YOUR_REQUEST_ID" \\
–subject "camera continues pulling back to reveal the vast city" \\
–duration 10<\/p>\n

OpenCode<\/p>\n

# Clone the repo and set the MUAPI_API_KEY env var
git clone https:\/\/github.com\/SamurAIGPT\/Generative-Media-Skills
export MUAPI_API_KEY=your_key_here<\/p>\n

# Skills auto-load from .opencode\/skills\/ when you run opencode in this directory
opencode<\/p>\n

\ud83e\udd16 MCP Server<\/p>\n

Run muapi as a Model Context Protocol server so Claude Desktop, Cursor, or any MCP-compatible agent can call generation tools directly \u2014 no shell scripts needed.<\/p>\n

muapi mcp serve<\/p>\n

Claude Desktop config ( ~\/Library\/Application Support\/Claude\/claude_desktop_config.json ):<\/p>\n

{
"mcpServers": {
"muapi": {
"command": "muapi",
"args": ["mcp", "serve"],
"env": { "MUAPI_API_KEY": "your-key-here" }
}
}
}<\/p>\n

This exposes 19 structured tools with full JSON Schema input\/output definitions:<\/p>\n

Tool
Description<\/p>\n

muapi_image_generate
Text-to-image (14 models)<\/p>\n

muapi_image_edit
Image-to-image editing (11 models)<\/p>\n

muapi_video_generate
Text-to-video (13 models)<\/p>\n

muapi_video_from_image
Image-to-video (16 models)<\/p>\n

muapi_audio_create
Music generation (Suno)<\/p>\n

muapi_audio_from_text
Sound effects (MMAudio)<\/p>\n

muapi_enhance_upscale
AI upscaling<\/p>\n

muapi_enhance_bg_remove
Background removal<\/p>\n

muapi_enhance_face_swap
Face swap image\/video<\/p>\n

muapi_enhance_ghibli
Ghibli style transfer<\/p>\n

muapi_edit_lipsync
Lip sync to audio<\/p>\n

muapi_edit_clipping
AI highlight extraction<\/p>\n

muapi_predict_result
Poll prediction status<\/p>\n

muapi_upload_file
Upload local file \u2192 URL<\/p>\n

muapi_keys_list
List API keys<\/p>\n

muapi_keys_create
Create API key<\/p>\n

muapi_keys_delete
Delete API key<\/p>\n

muapi_account_balance
Get credit balance<\/p>\n

muapi_account_topup
Add credits (Stripe checkout)<\/p>\n

\u26a1 Agentic Pipeline Examples<\/p>\n

# Submit async, capture request_id, poll when ready
REQUEST_ID=$(muapi video generate "a dog running on a beach" \\
–model kling-master –no-wait –output-json –jq '.request_id' | tr -d '"')<\/p>\n

# … do other work …<\/p>\n

muapi predict wait "$REQUEST_ID" –download .\/outputs<\/p>\n

# Pipe a prompt from another command
generate_prompt | muapi image generate – –model flux-dev<\/p>\n

# Chain: upload \u2192 edit \u2192 download
URL=$(muapi upload file .\/photo.jpg –output-json –jq '.url' | tr -d '"')
muapi image edit "make it look like a painting" –image "$URL" \\
–model flux-kontext-pro –download .\/outputs<\/p>\n

\ud83d\udcd6 Schema Reference<\/p>\n

This repository includes a streamlined schema_data.json that core scripts use at runtime to:<\/p>\n

Validate Model IDs : Ensures the requested model exists.<\/p>\n

Resolve Endpoints : Automatically maps model names to API endpoints.<\/p>\n

Check Parameters : Validates supported aspect_ratio , resolution , and duration values.<\/p>\n

Discover all available models via the CLI:<\/p>\n

muapi models list
muapi models list –category video –output-json<\/p>\n

\ud83d\udd27 Compatibility<\/p>\n

Optimized for the next generation of AI development environments:<\/p>\n

Claude Code \u2014 Direct terminal execution via tools + MCP server mode.<\/p>\n

Gemini CLI \/ Cursor \/ Windsurf \u2014 Seamless integration as local scripts.<\/p>\n

MCP \u2014 Full Model Context Protocol server with typed input\/output schemas.<\/p>\n

CI\/CD \u2014 –output-json , –jq , semantic exit codes for scripting.<\/p>\n

\ud83d\udcc4 License<\/p>\n

MIT \u00a9 2026<\/p>\n

\u8bd5\u8bd5\u8fd9\u6837\u505a<\/h2>\n