image-to-video
Transform static images into AI-animated videos
Also available from: inference-sh-9,doany-ai,agentspace-so
Creating videos from images requires selecting the right AI model for your use case. This skill intelligently routes your request to HappyHorse for portraits, Wan 2.7 for lip-sync, or Seedance for multi-modal animations.
Download the skill ZIP
Upload in Claude
Go to Settings → Capabilities → Skills → Upload skill
Toggle on and start using
Test it
Using "image-to-video". Animate this portrait photo with a slow dolly-in camera movement
Expected outcome:
A 5-second video clip of the portrait with smooth camera push-in, subtle facial expression changes, identity-preserved features, and soft natural lighting evolution.
Using "image-to-video". Create a talking-head video synced to this voiceover [attached audio file]
Expected outcome:
A 12-second vertical video of a spokesperson with precise lip-sync to the provided audio, medium close-up framing, and professional studio lighting.
Using "image-to-video". Make a brand video using this character image, this scene clip, and this voice sample
Expected outcome:
An 8-second video where the character from the image performs actions from the scene reference while speaking with the voice from the audio reference, maintaining visual consistency.
Security Audit
Low RiskAll 117 static findings are false positives. The skill is a markdown documentation file explaining how to use the RunComfy CLI for AI video generation. External command detections are example bash commands in fenced code blocks, not actual shell executions. Path traversal and filesystem detections are placeholder tokens and legitimate config paths. Network detections are service URLs, not data exfiltration. Security measures are properly documented with token storage at mode 0600, HTTPS transmission, and download size limits.
Medium Risk Issues (1)
Low Risk Issues (4)
Risk Factors
🌐 Network access (12)
⚙️ External commands (91)
Quality Score
What You Can Build
Animate product photos for e-commerce
Bring product images to life with smooth camera movements and subtle animations. HappyHorse preserves product geometry and packaging while adding cinematic motion.
Create talking-head videos with custom voiceovers
Generate spokesperson videos with lip-sync to your pre-recorded audio tracks. Perfect for localized marketing content with voice talent in any language.
Produce brand-consistent narrative videos
Combine a character image, scene from reference video, and voice from reference audio into a unified clip. Maintains identity consistency across brand assets.
Try These Prompts
Animate this portrait image with gentle camera movement around the subject's face, subtle breathing motion, and soft natural lighting.
Create a 360-style reveal animation for this product shot with smooth orbital camera movement, maintaining packaging and branding integrity.
Generate a spokesperson clip for 9:16 format that syncs lip movements to this audio track: [attach audio], with medium close-up framing and warm lighting.
Create a branded narrative clip using [subject image], [scene video reference], and [voice audio reference]. Subject from image 1 walks through the scene from video 1 with voice from audio 1.
Best Practices
- Lead prompts with motion verbs: drift, orbit, dolly in, reveal, blink. Front-load what is moving in the scene.
- Match duration to audio length when using lip-sync. The video will be silent past the audio duration if the clip is longer.
- Use explicit preservation keywords: identity-stable, packaging unchanged, background geometry stable when you need specific elements to remain consistent.
Avoid
- Do not restate the image content in prompts; the AI model sees the image. Focus tokens on what should change or animate.
- Do not request HappyHorse animation plus Wan-style lip-sync in a single call; these require separate model calls.
- Do not mix radically different aesthetics between reference image and reference video; output will drift from intended style.