Agent skill · creative production · heygen-com
heygen
heygen — an agent skill for coding agents.
Why this skill is useful
Adds specific commands and workflows for generating videos with AI avatars that the AI wouldn't reliably generate on its own.
What it needs
Requires heygen installed locally. Requires heygen account access. About 3k tokens when loaded. Last updated 2026-07-14. 382 stars on the source repository.
What this skill does
HeyGen Video Agent — NanoClaw Container Skill When to Use Use this skill when the user wants to create a video with an AI avatar presenter. Triggers: "make a video", "create a video message", "record a video", "avatar video", "talking head video", "video pitch", "video update". NOT for: image generation, audio-only TTS, video translation, or cinematic b-roll. Required Environment HEYGENAPIKEY — Get from https://app.heygen.com/settings?nav=API heygen CLI — install: curl -fsSL https://static.heygen.ai/cli/install.sh bash. Verify: heygen auth status. Steps Step 1: Discover Available Avatars avatar list returns avatar groups — each .id is a groupid, not an avatarid. The avatarid you pass to generation is a specific look: list looks with heygen avatar looks list --group-id <groupid> jq '.data[] {avatarid: .id, previewimageurl}' and pick a look's .id. If the user already has a specific look id, use it directly. Step 2: Find a Voice Pick a voiceid matching the desired language and tone. Step 3: Write the Script Write a spoken-word script for the avatar. Rules: Write for speech, not text. Short sentences. Natural pauses. 150 words per minute is the target pace. 30 seconds = ~75 words. 60 seconds = ~150 words. No stage directions. No markdown. Just what the avatar says. Step 4: Generate the Video With --wait, the CLI blocks until the video completes and emits the final status object. Without --wait, submission returns immediately — save the sessionid from stdout for manual polling. --timeout 10m is sized for short videos (≤60s). Bump to --timeout 45m for videos >60s or when targeting 2+ minute output — Video Agent generation routinely takes 20-45m for longer clips. Step 5: Poll for Completion (only without --wait) Poll every 15 seconds. Session status is one of thinking, waitingforinput, reviewing, generating, completed, failed — not a strict linear sequence (waitingforinput only occurs in chat mode). Terminal states are completed and failed. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills heygen-com/heygen