skills.rest
TFboy1
TFboy1COMMUNITY
Explore all skills by this author →
WebsiteGitHub

oh-my-minimaxh3-director

Automates script-to-video pipelines via MiniMax H3 storyboarding, ComfyUI generation, and JianYing assembly.

123|15|Updated Aug 13, 2026
Search Tags:#video generation#storyboard#comfyui#text-to-video#jianying#ai director#minimax h3
One-click install
npx skills add https://github.com/TFboy1/oh-my-minimaxh3-director --skill oh-my-minimaxh3-director-tfboy1
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: oh-my-minimaxh3-director
Source: https://github.com/TFboy1/oh-my-minimaxh3-director
Command: npx skills add https://github.com/TFboy1/oh-my-minimaxh3-director --skill oh-my-minimaxh3-director-tfboy1

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? Turning a written script into a finished video normally requires manual storyboarding, prompt engineering for video models, ComfyUI workflow configuration, batch job management, and video editing. This Skill automates that entire chain so a script or story text becomes an assembled JianYing draft with minimal manual steps. ## Core Features & Use Cases - Automatic storyboarding with continuity: Splits a script into 5-15 second segments with shot-to-shot continuity anchors, generates character four-view reference sheets, and writes MiniMax H3 prompts in official six-part English, WenWu Chinese director, or hybrid modes. - Workflow routing and batch generation: Routes each segment to Ref2VA, T2V, or I2V ComfyUI templates, validates parameters (duration 5-15s, steps 4-40), submits jobs with resumable monitoring, and supports Cloudflare tunnels or AutoDL cloud GPUs. - JianYing auto-assembly: Generates a JianYing draft with ordered clips and dissolve transitions, with optional MP4 export. - Use Case: Provide a game PV script and receive a storyboard with per-second timelines, confirmed generation parameters, downloaded video clips, and a ready-to-open JianYing draft. ## Quick Start Ask the AI to use $oh-my-minimaxh3-director to turn your script or story text into a finished video.

Frequently Asked Questions about oh-my-minimaxh3-director

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

FAQPage Schema
How do I turn a script into a video with MiniMax H3?▼

Provide your script or story text and the Skill generates a storyboard with shot continuity, writes H3 prompts, builds ComfyUI workflows, submits generation jobs, and assembles clips into a JianYing draft. You confirm parameters in a summary table before rendering starts.

What hardware do I need to run MiniMax H3 video generation?▼

You need an NVIDIA GPU with CUDA; 16-24GB VRAM is recommended for 0.4-0.6MP 10-second clips, while under 12GB is not recommended locally. The check_hardware.py script evaluates your machine, and cloud GPUs via AutoDL are supported as an alternative.

Can I use my own ComfyUI workflows instead of the built-in templates?▼

Yes, the scan_workflows.py script scans your existing workflows and presents candidates for you to choose. UI-format workflows must first be converted to API format with convert_ui_workflow.py; the Skill never silently rewrites your workflow files.

Why does my ComfyUI submission return node_errors?▼

node_errors usually means the remote ComfyUI is missing a custom node or the model filename does not match. Read the class_type in the error message, install the missing node, or check available model names via the /object_info endpoint.

Does the pipeline support unattended batch rendering?▼

Yes, an unattended mode asks for a shutdown time before each batch and runs monitor_resources.py to watch GPU VRAM and system memory with 90% warning and 95% stop thresholds. Cloud AutoDL instances can be powered off via API at the agreed time.

What are the limitations of the generated JianYing draft?▼

The draft contains ordered clips with 0.3-second dissolve transitions but no subtitles or background music. Automatic MP4 export requires Windows with JianYing 5.9 or earlier and only runs when you explicitly request it.

Related Skills

Explore More Design & Creative Skills→
artokunartokun
715

model-registry

Maps curated model download URLs to ComfyUI target directories by model family.

Community
Intermediate
ai-service-incubatorai-service-incubator
1

generate-video

Generate 1080x1920 MP4 short-form videos from scripts with subtitles, images, and music.

Official
Intermediate
i-bebsii-bebsi

comfyui

Generate images, video, and audio by running ComfyUI workflows via REST and WebSocket APIs.

Community
Advanced
atomicrajatatomicrajat
2

vss-summarize-video

Summarize recorded video clips via the LVS microservice with a VLM fallback.

Community
Advanced