heygen-avatar

name	heygen-avatar
description	Create a persistent HeyGen avatar that looks and sounds like a specific person — the user, the agent, or any named character — powered by HeyGen Avatar V technology. Upload a photo → HeyGen builds a digital twin → reuse across unlimited videos. Use when: (1) someone wants to appear in a video as themselves ("I want my face in a video", "create my HeyGen avatar", "build a digital twin of me"), (2) setting up a HeyGen identity before making videos or sending video messages — the correct FIRST step for new users, (3) "create my avatar", "design an avatar", "give me a consistent look across my videos", "bring yourself to life", "set up my identity on HeyGen", "set up my HeyGen identity", "get started with HeyGen", "help me get started with AI video". Chain signal: when the user says both an identity/avatar action AND a video action in the same request ("design an avatar AND make a video", "set up my identity THEN create a video", "design a presenter AND immediately record"), run heygen-avatar first, then heygen-video. Returns avatar_id + voice_id — pass directly to heygen-video to create HeyGen videos. NOT for: generating videos (use heygen-video), translating videos, or TTS-only tasks.
argument-hint	[photo_url_or_description]

HeyGen Avatar Designer

Create and manage HeyGen avatars for anyone: the agent, the user, or named characters. Handles identity extraction, avatar generation, voice selection, and saves everything to AVATAR-<NAME>.md for consistent reuse.

Before You Start (Claude Code only)

Try to read SOUL.md from the workspace root.

• Found → OpenClaw environment. Skip this section entirely and go straight to Phase 0.
• Not found → Claude Code environment. Say this before anything else:

First, fetch the user's existing HeyGen avatars: GET https://api.heygen.com/v3/avatars (no query params — the endpoint returns private avatars for the authenticated key). Parse the data array.

⚠️ AVATAR file caveat: Ignore any AVATAR-*.md files found in the workspace that belong to a different person or agent (e.g., AVATAR-Eve.md when creating an avatar for Claude). Only use an AVATAR file if its name matches the subject you're creating for right now.

If the user has existing avatars (non-empty data array), present them as numbered options and ask which to use or whether to create a new one. Communicate in user_language.

If the user has no existing avatars (empty data), tell them none were found and offer to create one with a few quick questions. Mention the OpenClaw SOUL.md shortcut for future reference. Communicate in user_language.

Wait for their answer before proceeding.

Required: HEYGEN_API_KEY. Resolved in order: (1) env var, (2) ~/.heygen/config file (source ~/.heygen/config 2>/dev/null). If neither found, tell the user to run ./setup or export HEYGEN_API_KEY=<key>. API: v3 only. Base: https://api.heygen.com. Never use v1 or v2 endpoints.

Required headers on every API request — no exceptions:

X-Api-Key: $HEYGEN_API_KEY
User-Agent: HeyGen-Stack/1.2.7 (OpenClaw; heygen-stack)
X-HeyGen-Source: openclaw-skill

Docs-first rule: Before calling any endpoint you're unsure about:

• Index: GET https://developers.heygen.com/llms.txt — full sitemap
• Any page: Append .md to the URL for clean markdown
• Read the spec, THEN build your request. Never guess field names.

Avatar File Convention

Every avatar gets one file: AVATAR-<NAME>.md at the workspace root.

AVATAR-EVE.md      ← agent
AVATAR-KEN.md      ← user
AVATAR-CLEO.md     ← named character

Format:

# Avatar: <Name>

## Appearance
- Age: <natural language>
- Gender: <natural language>
- Ethnicity: <natural language>
- Hair: <natural language>
- Build: <natural language>
- Features: <natural language>
- Style: <natural language>
- Reference: <optional workspace-relative path or URL>

## Voice
- Tone: <natural language>
- Accent: <natural language>
- Energy: <natural language>
- Think: 


 Group ID: 
 Voice ID: 
 Voice Name: 
 Voice Designed: 
 Voice Seed: 
 Looks: landscape=, portrait=, square=
 Last Synced: 

⚠️ lookid at runtime via GET /v3/avatars/looks?groupid as the primary avatar reference.

Top sections (Appearance, Voice) are portable natural language. Any platform can use them. HeyGen section is runtime config with API IDs. Skills read this to make API calls.

Skill Announcement

Start every invocation with:

🎭 Using: heygen-avatar — creating an avatar for [name]

Workflow

Phase 0 — Who Are We Creating?

Determine the target identity:

1. Agent — user says "create your avatar", "bring yourself to life" → read IDENTITY.md for name, then check AVATAR-<NAME>.md. If IDENTITY.md is not found (Claude Code environment), walk the user through designing from scratch with a few quick questions about appearance and voice.
2. User — user says "create my avatar", "make me an avatar" → ask for their name, check AVATAR-<NAME>.md
3. Named character — user says "create an avatar called Cleo" → check AVATAR-CLEO.md

If the AVATAR file exists and has a HeyGen section filled in:

"You already have an avatar set up. Want to add a new look, update it, or start fresh?"

If the AVATAR file exists but HeyGen section is empty: proceed to Reference Photo Nudge. If no AVATAR file exists: proceed to Phase 1.

Reference Photo Nudge (First-Time Only)

Before generating anything, ask if they have a reference image. Photo avatars produce significantly better face consistency across videos than prompt-generated ones.

Ask if they have a reference photo, explaining that a headshot or clear face photo gives much better results than text-only generation. Offer to skip for prompt-based creation. Communicate in user_language.

This applies to ALL targets (agent, user, named character). For agents, check if a reference photo path already exists in the AVATAR file's Appearance section or in IDENTITY.md before asking.

• Photo provided → upload via POST /v3/assets, then use Type B (photo) creation in Phase 2
• Skip → use Type A (prompt) creation in Phase 2

Phase 1 — Identity Extraction

For the agent: Try to read SOUL.md, IDENTITY.md, and existing AVATAR-<NAME>.md from the workspace. If found, extract appearance and voice traits automatically. If not found (e.g. Claude Code environment), skip to conversational onboarding — ask the user to describe the agent's appearance and voice instead.

For users/named characters: Conversational onboarding. Ask naturally about their appearance (age, hair, general vibe) and voice (calm, energetic, accent). Not as a form — be conversational. Communicate in user_language.

Write AVATAR-<NAME>.md with the Appearance and Voice sections filled in. Leave HeyGen section empty.

Then proceed to the Reference Photo Nudge before Phase 2.

Phase 2 — Avatar Creation

API: POST https://api.heygen.com/v3/avatars

Two modes via the same endpoint:

Mode 1 — New character (omit avatar_group_id): Creates a brand new character with its own group.

Mode 2 — New look (include avatar_group_id): Adds a variation to an existing character. Read the Group ID from the AVATAR file.

Two creation types:

Type A — From prompt (AI-generated appearance):

{
  "type": "prompt",
  "name": "<name>",
  "prompt": "<appearance prompt, max 1000 chars>",
  "avatar_group_id": "<optional — Mode 2 only>"
}

Prompt limit is 1000 characters. Be descriptive — include style, features, expression, lighting. The API spec says 200 but the actual enforced limit is 1000.

Type B — From reference image:

{
  "type": "photo",
  "name": "<name>",
  "file": { "type": "url", "url": "https://..." },
  "avatar_group_id": "<optional — Mode 2 only>"
}

File options for Type B:

• { "type": "url", "url": "https://..." } — public image URL
• { "type": "asset_id", "asset_id": "<id>" } — from asset upload
• { "type": "base64", "media_type": "image/png", "data": "<base64>" } — inline

To upload a local file first:

POST https://api.heygen.com/v3/assets
Content-Type: multipart/form-data
Body: file=@<photo_path>

Response: Returns avatar_item.id (look ID) and avatar_item.group_id (character identity).

Map identity fields to HeyGen enums for the prompt:

• age: Young Adult | Early Middle Age | Late Middle Age | Senior | Unspecified
• gender: Man | Woman | Unspecified
• ethnicity: White | Black | Asian American | East Asian | South East Asian | South Asian | Middle Eastern | Pacific | Hispanic | Unspecified
• style: Realistic | Pixar | Cinematic | Vintage | Noir | Cyberpunk | Unspecified
• orientation: square | horizontal | vertical
• pose: half_body | close_up | full_body

Show the prompt to the user before creating:

Appearance: "[prompt]" Settings: Young Adult | Woman | East Asian | Realistic Look good? (yes / adjust / completely different)

⛔ STOP. Wait for the user to approve or adjust. Do NOT call the avatar creation API until the user confirms.

Phase 3 — Voice

Two paths: Design (describe what you want, get matched voices) or Browse (filter the catalog manually).

Ask whether they want voice design (describe what they want) or catalog browsing. Communicate in user_language.

Default to Design if the AVATAR file has a Voice section with personality traits.

Path A — Voice Design (preferred)

Find matching voices via semantic search using the Voice section from the AVATAR file. This searches HeyGen's full voice library. No new voices are generated and no quota is consumed.

Language matching: The voice design prompt should specify the target language from user_language. Example for Japanese: "A calm, warm female voice. Professional but approachable. Japanese speaker." This ensures semantic search returns voices in the correct language.

POST https://api.heygen.com/v3/voices
{
  "prompt": "<built from AVATAR Voice section: tone, accent, energy, personality. Include target language.>",
  "seed": 0
}

Returns 3 voice options per seed. Present all 3 with inline audio previews:

• Download each preview_audio_url: curl -sL "<url>" -o /tmp/voice-design-<n>.mp3
• Send as audio attachment: message(action:send, media:"/tmp/voice-design-<n>.mp3", caption:"Option <n>: <voice_name> — <gender>, <language>") so it plays inline in Telegram/Discord
• After all previews sent, present selection buttons

⛔ STOP. Wait for the user to pick a voice via buttons or text. Do NOT select a voice yourself or proceed to Phase 4 until the user explicitly chooses.

If none match:

"None of these hitting right? I can try a different set (same description, different variations) or you can tweak the description."

Increment seed and call again. Different seeds give completely different voice options from the same prompt.

• Clean up /tmp files after user picks

Path B — Voice Browse (fallback)

Browse HeyGen's existing voice library:

GET https://api.heygen.com/v3/voices

1. Read the Voice section from the AVATAR file
2. Filter by gender and language
3. Pick top 3 candidates based on personality match
4. Present with inline audio previews (same download + send pattern as Path A)
5. ⛔ STOP. Wait for the user to pick. Do NOT auto-select.

Phase 4 — Save to AVATAR File

Update the HeyGen section of AVATAR-<NAME>.md to match the canonical format:

## HeyGen
- Group ID: <avatar_item.group_id — THE stable reference, never changes>
- Voice ID: <chosen voice_id>
- Voice Name: <voice name>
- Voice Designed: <true if custom-designed, false if picked from catalog>
- Voice Seed: <seed value used, if designed>
- Looks: <orientation>=<avatar_item.id> (e.g., landscape=<look_id>, portrait=<look_id>)
- Last Synced: <ISO timestamp>

⚠️ look_ids are ephemeral — always resolve fresh from group_id at runtime via GET /v3/avatars/looks?group_id=<id>. Never hardcode look_id as the primary avatar reference.

Confirm the avatar is saved and that other skills (like heygen-video) will pick it up automatically. Communicate in user_language.

Phase 5 — Test (Optional)

If the user wants to see their avatar in action:

POST https://api.heygen.com/v3/video-agents
{
  "avatar_id": "<avatar_id>",
  "voice_id": "<voice_id>",
  "prompt": "<short greeting in the video language>"
}

Generate a natural greeting in the video language (from user_language). Examples: English "Hi, I'm [name]. Nice to meet you!", Japanese "[name]です。はじめまして！", Spanish "Hola, soy [name]. ¡Mucho gusto!", Korean "안녕하세요, [name]입니다. 만나서 반갑습니다!"

Iteration Flow

When the user wants to refine:

• "Adjust the prompt" → Mode 2 with existing group_id (keeps the character, adds a new look). Only Mode 1 if they say "start completely over."
• "Add a new look" / "different outfit" → Mode 2 with existing group_id. Add to Looks in AVATAR file.
• "Try a different voice" → back to Phase 3
• "Start completely over" → Mode 1, new character. Overwrite HeyGen section.

Default to Mode 2 (new look under same group). Only create a new group when the user explicitly wants a different character identity. This keeps the account clean and makes looks reusable across skills.

Each iteration updates the AVATAR file. The file is always the source of truth.

UX Rules

Be interactive at decision points, silent everywhere else.

• At checkpoints (avatar approval, voice selection): STOP and wait for user input. Never auto-select.
• Between checkpoints: work silently. Do not narrate your reasoning, list parameters, or explain what you're about to do.
• After the user picks a voice: save the file and confirm. One message. Don't recap the journey.

Video Producer Integration

heygen-video reads AVATAR files for group_id and voice_id:

• "Make a video with Eve" → reads AVATAR-EVE.md → gets Group ID + Voice ID → resolves fresh look_id at runtime
• "Make a video with Ken" → reads AVATAR-KEN.md
• No AVATAR file found → falls back to stock avatars or asks user

Error Handling

• Missing SOUL.md/IDENTITY.md → conversational onboarding, write AVATAR file from answers
• API fails → retry once, then ask user to check API key
• Voice match poor → show all available voices, let user browse
• Asset upload fails → skip reference image, try prompt-only creation
• Existing avatar file with stale HeyGen IDs → offer to regenerate or keep