create-video

Generate videos from text prompts using the HeyGen Video Agent API.

Updated Jul 16, 2026
One-click install
npx skills add https://github.com/X-manist/Cohmira --skill create-video-x-manist
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-video
Source: https://github.com/X-manist/Cohmira/tree/main/src/builtin-plugins/openmontage/.agents/skills/create-video
Command: npx skills add https://github.com/X-manist/Cohmira --skill create-video-x-manist

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning a rough idea into a finished video normally requires scripting, avatar selection, visual design, voiceover, and editing. This Skill automates that entire pipeline by sending a single optimized prompt to HeyGen's Video Agent, which handles script writing, avatar selection, visuals, voiceover, pacing, and captions automatically. ## Core Features & Use Cases - Prompt-to-Video Generation: Describe the video you want and generate a complete draft via the POST /v1/video_agent/generate endpoint or the mcp__heygen__generate_video_agent MCP tool. - Prompt Optimization Framework: Built-in guidance for writing production-quality prompts with named visual styles, scene-by-scene structure, layered motion graphics, and critical on-screen text. - Status Polling and Download: Poll video status with mcp__heygen__get_video or GET /v2/videos/{video_id} and retrieve download URLs once rendering completes. - Use Case: A marketing team member says "make me a 60-second product demo video" and receives a rendered MP4 with avatar narration, motion graphics, and captions without touching a video editor. ## Quick Start Ask the AI to create a 60-second landscape video about your product using the create-video skill, making sure the HEYGEN_API_KEY environment variable is set first.

Frequently Asked Questions about create-video

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

FAQPage Schema
How do I generate a video from a text prompt with HeyGen?

Send an optimized prompt to the Video Agent endpoint POST /v1/video_agent/generate with your X-Api-Key header, or use the mcp__heygen__generate_video_agent MCP tool. The API returns a video_id that you poll until the status becomes completed.

What is the difference between HeyGen Video Agent and the standard video API?

Video Agent generates a full video from one prompt, automatically handling script, avatar, visuals, and captions. The standard v2/video/generate API gives precise control over avatars, exact scripts, per-scene voices, and multi-scene composition.

How long does HeyGen video generation take?

Generation typically takes 5 to 15 minutes, but can exceed 20 minutes during peak load or for longer scripts. Set polling timeouts of 15 to 20 minutes and consider webhooks or resumable status checks for long videos.

Can I control video duration and orientation in HeyGen Video Agent?

Yes. Pass a config object with duration_sec (5 to 300 seconds), orientation (portrait or landscape), and optionally avatar_id to lock a specific avatar. Duration is approximate rather than exact.

Why does my HeyGen Video Agent output look empty or low quality?

Poor results usually come from weak prompts. Avoid layout coordinates and B-roll scenes under 5 seconds, and instead use named visual styles, layered motion descriptions with action verbs, scene rotation, and voiceover on every scene.

When should I not use prompt-based video generation?

Avoid it when you need exact script wording, a specific avatar saying precise lines, transparent WebM output, or multi-scene composition with per-scene backgrounds. Use the avatar-video skill or standard v2 API for that level of control.