Index of every page on this site: llms.txt
All skills
openmontageinc avatar

/avatar-video

@2e003e6
by openmontageincopenmontageinc/openmontage82k stars
11,135
Run it yourself

Create AI avatar videos with precise control over avatars, voices, scripts, scenes, and backgrounds using HeyGen's v2 API. Use when: (1) Choosing a specific avatar and voice for a video, (2) Writing exact scripts for an avatar to speak, (3) Building multi-scene videos with different backgrounds per scene, (4) Creating transparent WebM videos for compositing, (5) Using talking photos as video presenters, (6) Integrating HeyGen avatars with Remotion, (7) Batch video generation with exact specs, (8) Brand-consistent production videos with precise control.

Use this Skill: https://skilld.dev/gh/openmontageinc/openmontage/avatar-video

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈143 tokens for metadata: the name and description. ≈1.5k when used: this file. ≈44k more on demand in 15 files.

Description uses 7.2% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈573 characters, or 7.2%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Avatar Video

Create AI avatar videos with full control over avatars, voices, scripts, scenes, and backgrounds. Build single or multi-scene videos with exact configuration using HeyGen's /v2/video/generate API.

Authentication

All requests require the X-Api-Key header. Set the HEYGEN_API_KEY environment variable.

curl -X GET "https://api.heygen.com/v2/avatars" \
  -H "X-Api-Key: $HEYGEN_API_KEY"

Tool Selection

If HeyGen MCP tools are available (mcp__heygen__*), prefer them over direct HTTP API calls — they handle authentication and request formatting automatically.

Task MCP Tool Fallback (Direct API)
Check video status / get URL mcp__heygen__get_video GET /v2/videos/{video_id}
List account videos mcp__heygen__list_videos GET /v2/videos
Delete a video mcp__heygen__delete_video DELETE /v2/videos/{video_id}

Video generation (POST /v2/video/generate) and avatar/voice listing are done via direct API calls — see reference files below.

Default Workflow

  1. List avatars — GET /v2/avatars → pick an avatar, preview it, note avatar_id and default_voice_id. See avatars.md
  2. List voices (if needed) — GET /v2/voices → pick a voice matching the avatar's gender/language. See voices.md
  3. Write the script — Structure scenes with one concept each. See scripts.md
  4. Generate the video — POST /v2/video/generate with avatar, voice, script, and background per scene. See video-generation.md
  5. Poll for completion — GET /v2/videos/{video_id} until status is completed. See video-status.md

Quick Reference

Task Read
List and preview avatars avatars.md
List and select voices voices.md
Write and structure scripts scripts.md
Generate video (single or multi-scene) video-generation.md
Add custom backgrounds backgrounds.md
Add captions / subtitles captions.md
Add text overlays text-overlays.md
Create transparent WebM video video-generation.md (WebM section)
Use templates templates.md
Create avatar from photo photo-avatars.md
Check video status / download video-status.md
Upload assets (images, audio) assets.md
Use with Remotion remotion-integration.md
Set up webhooks webhooks.md

When to Use This Skill vs Create Video

This skill is for precise control — you choose the avatar, write the exact script, configure each scene.

If the user just wants to describe a video idea and let AI handle the rest (script, avatar, visuals), use the create-video skill instead.

User Says Create Video Skill This Skill
"Make me a video about X" ✓
"Create a product demo" ✓
"I want avatar Y to say exactly Z" ✓
"Multi-scene video with different backgrounds" ✓
"Transparent WebM for compositing" ✓
"Use this specific voice for my script" ✓
"Batch generate videos with exact specs" ✓

Reference Files

Core Video Creation

Video Customization

Advanced Features

Integration

Foundation

Best Practices

  1. Preview avatars before generating — Download preview_image_url so the user can see the avatar before committing
  2. Use avatar's default voice — Most avatars have a default_voice_id pre-matched for natural results
  3. Fallback: match gender manually — If no default voice, ensure avatar and voice genders match
  4. Use test mode for development — Set test: true to avoid consuming credits (output will be watermarked)
  5. Set generous timeouts — Video generation often takes 5-15 minutes, sometimes longer
  6. Validate inputs — Check avatar and voice IDs exist before generating

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 alert12d3 checks · Risk CRITICAL
  • Gen Agent Trust Hub12d

    The skill provides comprehensive instructions for generating AI avatar videos using the HeyGen API. It contains an indirect prompt injection surface as it handles untrusted user-provided scripts and assets without specific boundary markers or sanitization. Automated security scan flags were reviewed, but no malicious patterns, obfuscation, or exfiltration indicators were identified in the instruction files.

  • Socket12d

    No alerts

  • Snyk12d

    Risk: LOW · No issues

Signed by skilld at 2e003e6. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 4 days ago.

Activeupdated 8 months ago
MCP servers
heygen
homepage
https://docs.heygen.com/reference/create-a-video
All 1 allowed tools
mcp__heygen__*
Other metadata
metadata
{
  "openclaw": {
    "requires": {
      "env": [
        "HEYGEN_API_KEY"
      ]
    },
    "primaryEnv": "HEYGEN_API_KEY"
  }
}

README badge

README badge for openmontageinc/openmontage/avatar-video