skills.rest
TeamWiseFlow
TeamWiseFlowOFFICIAL
Explore all skills by this author →
GitHub

aigc-video-gen

Generate short video clips with synchronized audio via DashScope, Volcengine, or MiniMax APIs.

8.5k|1.4k|Updated Apr 24, 2024
Search Tags:#music generation#video generation#text-to-video#dashscope#AIGC#Seedance#MiniMax
One-click install
npx skills add https://github.com/TeamWiseFlow/wiseflow --skill aigc-video-gen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aigc-video-gen
Source: https://github.com/TeamWiseFlow/wiseflow/tree/main/skills/aigc-video-gen
Command: npx skills add https://github.com/TeamWiseFlow/wiseflow --skill aigc-video-gen

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Producing short video clips for social media content pipelines requires calling multiple AIGC video providers with different APIs, model chains, and constraints. This Skill unifies Aliyun DashScope (happyhorse/wan2.7), Volcengine Ark (Seedance), and MiniMax Hailuo (H3) behind one command, handling platform detection, model fallback chains, and async task polling automatically.

Core Features & Use Cases

  • Multi-provider video generation: Supports t2v (text-to-video), i2v (image-to-video), and r2v (reference-to-video) modes across three platforms, with automatic platform selection based on configured API keys and per-mode model candidate chains with fallback.
  • Audio-visual co-generation: All supported models generate synchronized voiceover, BGM, and ambient sound in one pass; MiniMax additionally offers a standalone background music generation subcommand (music-3.0).
  • Segment continuity tooling: The --prev-segment flag extracts the last frame of the previous clip via ffmpeg to align first frames across multi-segment character stories.
  • Use Case: A content producer building a 45-second vertical short video splits it into five 9-second clips, generates each with r2v using a character reference image, chains them with --prev-segment for visual continuity, and generates a 30-second BGM track via the MiniMax music subcommand.

Quick Start

Run aigc-video-gen with a Chinese prompt describing visuals and audio, a duration of 3-15 seconds, a 9:16 ratio, and an output path under output_videos/ to generate a video clip with synchronized sound.

Frequently Asked Questions about aigc-video-gen

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate AI video clips with synchronized audio?▼

Run the aigc-video-gen command with a --prompt describing both visuals and audio, plus --duration, --ratio, and --output. The wrapper auto-detects the provider from environment variables and submits an async generation task, polling until the MP4 is ready.

What is the difference between t2v, i2v, and r2v video generation modes?▼

t2v generates video from text only, i2v uses a first-frame image via --image, and r2v uses a reference image via --ref-image for character consistency. The mode is inferred automatically from which flags you pass.

Which API keys are required for MiniMax, Volcengine, and DashScope video generation?▼

MiniMax requires MINIMAX_API_KEY, Volcengine Seedance requires AWK_GEN_KEY (not ARK_API_KEY), and Aliyun DashScope requires MODELSTUDIO_API_KEY or DASHSCOPE_API_KEY. If none are set, the script exits with code 2 and suggests stock footage alternatives.

Can I generate background music with MiniMax Hailuo?▼

Yes, the music subcommand calls MiniMax music-3.0 to generate MP3 background music from a text prompt describing style, mood, and instruments. It requires MINIMAX_API_KEY and is only available on the MiniMax platform.

What is the maximum video duration for AI-generated clips?▼

Clips are limited to 3-15 seconds on DashScope, 2-15 seconds on Volcengine, and 4-15 seconds on MiniMax H3. Longer content must be split into multiple segments, and --prev-segment can align first frames between consecutive clips.

Why does video generation fail when no API key is configured?▼

The wrapper checks MINIMAX_API_KEY, AWK_GEN_KEY, and MODELSTUDIO_API_KEY/DASHSCOPE_API_KEY in order. If none exist, it exits with code 2 and instructs the agent to fall back to pexels-footage or pixabay-footage stock footage skills instead.

Related Skills

Explore More Content & Communication Skills→
akrindevakrindev

scene-detector

Detect scene changes and shot boundaries in videos using PySceneDetect.

Community
Intermediate
icoolworldicoolworld

distribution-operator

Generate platform-aware publishing variants for titles, covers, hashtags, and posting times.

Community
Advanced
RhythrosaLabsRhythrosaLabs
1

video_production

Turn input images into promotional videos with AI motion, voiceover, and music.

Community
Advanced
video-production-buddyvideo-production-buddy
602

stock-sourcing

Source and download stock images and video clips with licensing metadata.

Official
Intermediate