A collection of skills for working with the HeyGen AI video creation API, designed for use with Claude Code and other skills-compatible agents.
| Skill | Description |
|---|---|
| create-video | Create videos from a text prompt — describe what you want and AI handles the rest (Video Agent API) |
| avatar-video | Build videos with precise control over avatars, voices, scripts, scenes, and backgrounds (v2 API) |
| ai-video-gen | Generate AI videos from text prompts with multiple provider options (VEO, Kling, Sora, Runway) via the Workflow Gateway |
| faceswap | Swap a face from a source image into a target video using GPU-accelerated AI via the Workflow Gateway |
| text-to-speech | Generate standalone speech audio from text using HeyGen's Starfish TTS model with voice, speed, pitch, and emotion control |
| video-translate | Translate and dub existing videos into 12+ languages with lip-sync, voice cloning, and multi-speaker support |
| visual-style | Create, extract, and apply portable visual design systems (visual-style.md) across HeyGen, slides, Figma, and more |
| video-download | Download video and audio from YouTube and 1000+ sites using yt-dlp |
| video-edit | Edit videos locally using ffmpeg — trim, concat, resize, speed, overlay, extract audio, compress, and convert |
| video-understand | Understand video content locally using ffmpeg frame extraction and Whisper transcription — no API keys needed |
| heygen | (Deprecated) Legacy combined skill — use create-video or avatar-video instead |
/plugin marketplace add heygen-com/skills
/plugin install heygen@heygen-skillsInstall using the skills CLI:
# Install all skills globally
npx skills add heygen-com/skills -a claude-code -g
# Or install to current project only
npx skills add heygen-com/skills -a claude-code
# List available skills first
npx skills add heygen-com/skills --list
# Install a specific skill only
npx skills add heygen-com/skills --skill create-videoThis works with Claude Code, Cursor, Codex, and other agents.
Clone and symlink to your Claude skills directory:
# Clone the repository
git clone https://github.com/heygen-com/skills.git
# Symlink all skills to personal skills directory (available in all projects)
for skill in skills/skills/*/; do
ln -s "$(pwd)/$skill" ~/.claude/skills/$(basename "$skill")
done
# OR symlink to project skills (available in current project only)
mkdir -p .claude/skills
for skill in skills/skills/*/; do
ln -s "$(pwd)/$skill" .claude/skills/$(basename "$skill")
doneThe skills should appear when Claude Code loads. You can verify by asking Claude about HeyGen APIs.
| Task | Skill |
|---|---|
| Generate video from a prompt | create-video |
| Generate video with precise scene control | avatar-video |
| List avatars and voices | avatar-video |
| Generate AI video from a text prompt | ai-video-gen |
| Swap a face into a video | faceswap |
| Generate speech audio from text | text-to-speech |
| List TTS voices | text-to-speech |
| Translate/dub an existing video | video-translate |
| Batch translate to multiple languages | video-translate |
| Create a visual design system | visual-style |
| Extract a visual style from a website/video/PDF | visual-style |
| Apply a visual style to HeyGen/slides/Figma | visual-style |
| Download a video from YouTube or other sites | video-download |
| Extract audio from a video URL | video-download |
| Trim, resize, or compress a video | video-edit |
| Concatenate video clips | video-edit |
| Extract or replace audio in a video | video-edit |
| Understand / transcribe a local video | video-understand |
| Extract key frames from a video | video-understand |
"Create a 60-second product demo video"
"I need avatar Josh to read this exact script with a blue background"
"Generate a TTS audio file with an excited female voice"
"Translate this YouTube video to Spanish and French"
"Build a 3-scene video: intro, feature demo, and CTA with different backgrounds"
"Use the prompt optimizer to create a scene-by-scene script"
"Create a visual-style.md for a retro 80s neon aesthetic"
"Extract a visual style from https://stripe.com"
"Apply the Swiss International Style to a HeyGen video about Q4 results"
All skills use the HeyGen API:
- Base URL:
https://api.heygen.com - Authentication:
X-Api-Keyheader - API Versions: v1, v2 (varies by endpoint)
| Endpoint | Skill | Purpose |
|---|---|---|
POST /v1/video_agent/generate |
create-video | One-shot prompt-to-video |
POST /v2/video/generate |
avatar-video | Precise multi-scene video generation |
GET /v2/videos/{video_id} |
create-video, avatar-video | Get video details and status |
GET /v2/videos |
create-video, avatar-video | List account videos |
DELETE /v2/videos/{video_id} |
create-video, avatar-video | Delete a video |
GET /v2/avatars |
avatar-video | List available avatars |
GET /v2/voices |
avatar-video | List available voices |
POST /v1/audio/text_to_speech |
text-to-speech | Generate speech audio |
GET /v1/audio/voices |
text-to-speech | List TTS-compatible voices |
POST /v2/video_translate |
video-translate | Start video translation |
POST /v1/workflows/executions |
ai-video-gen, faceswap | Execute a workflow (video generation, face swap) |
GET /v1/workflows/executions/{id} |
ai-video-gen, faceswap | Check workflow execution status |
- HeyGen API key (get one at HeyGen Developer Portal)
- Claude Code CLI (or any skills-compatible agent)
To add or update skills:
- Edit the relevant
SKILL.mdor reference files inskills/<skill-name>/ - Include curl, TypeScript, and Python examples where applicable
- Test that the skill loads correctly
- Each skill should be self-contained — no cross-skill dependencies
MIT