Who it is forA developer using Claude Code or Cursor who wants visuals for a project without switching tools.
The problemCoding agents write code well but have no built-in way to generate images, video or audio.
The outcomeYour agent can call the higgsfield-generate skill to make images, video, 3D and audio from a prompt.
What you need first
- An AI coding agent that loads Markdown-based skills, such as Claude Code, Cursor or Codex
- A Higgsfield AI account; skill setup handles the Higgsfield CLI install and auth
- For gh skill install: GitHub CLI v2.90 or newer
Steps
- Run
npx skills add higgsfield-ai/skills(the recommended cross-agent install) - Let the skill setup install the Higgsfield CLI and sign in
- Ask your agent to use the higgsfield-generate skill with your prompt
- For image to video, the README suggests
seedance_2_5with--mode omni_reference --start-image
Prompt for your agent
Install higgsfield-ai/skills with `npx skills add higgsfield-ai/skills`, then use the higgsfield-generate skill to make <image or video> of <subject>. Show me the result before generating variations.
Caveat. Generation runs through Higgsfield AI, so an account is needed and usage may be billed by them.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for higgsfield-ai/skills and "Add image and video generation to your coding agent"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer