Video Extend

Takes an existing Veo 3-1 video clip and continues it with consistent motion, lighting, and character identity. You pick a source video URL, write a prompt describing what happens next (not what's already there), and the model produces the continuation. Standard quality for final delivery, fast variant for iteration. The prompting guidance is solid: anchor the camera explicitly, keep each extend to one story beat, and chain consecutive extends if you need a longer sequence. Identity drift accumulates after 3-4 chained generations, so keep individual extends short. Uses the RunComfy CLI with the `google-deepmind/veo-3-1/extend-video` endpoint. Useful when you hit the per-call duration cap or want to build a multi-shot narrative from a single seed clip.

skills.sh installs: 117.5K
GitHub stars: 11
last updated: Jun 3, 2026

View on GitHub

Install

$ npx -y skills add agentspace-so/runcomfy-agent-skills --skill video-extend --agent claude-code

Installs into .claude/skills of the current project.

Files

Video Extend

Continue an existing video clip past its per-call duration cap, or chain a narrative shot-by-shot from a single seed. This skill routes to Google Veo 3-1's extend-video endpoints and ships the documented prompting patterns + the exact runcomfy run invoke.

runcomfy.com · Veo 3-1 extend-video · CLI docs

Powered by the RunComfy CLI

# 1. Install (see runcomfy-cli skill for details)
npm i -g @runcomfy/cli      # or:  npx -y @runcomfy/cli --version

# 2. Sign in
runcomfy login              # or in CI: export RUNCOMFY_TOKEN=<token>

# 3. Extend
runcomfy run google-deepmind/veo-3-1/extend-video \
  --input '{"video_url": "https://...", "prompt": "..."}' \
  --output-dir ./out

CLI deep dive: runcomfy-cli skill.