Skip to content
Kitbay

higgsfield-faceless-video

MITUpdated on Nov 5, 20262,026GitHub starsIncludes scripts

A skill from Higgsfield's own collection that walks an agent through producing a narrated video with no on-screen presenter: an illustrated explainer, a documentary, a picture story or a sung children's video, or an edit of one it made before.

What the skill teaches

A skill from Higgsfield's own collection that walks an agent through producing a narrated video with no on-screen presenter: an illustrated explainer, a documentary, a picture story or a sung children's video, or an edit of one it made before. The agent first settles topic, length, aspect ratio, language, visual style, voice, music and subtitles, then uses the Higgsfield command-line tool to generate scenes and narration and joins them locally with ffmpeg through an included shell script. Before any paid generation it checks the account, the required programs and the cost estimate. It needs the Higgsfield CLI signed in, ffmpeg, ffprobe and Python 3.

Capabilities

  • Plan and generate narrated explainer, documentary or story videos
  • Make sung children's music videos
  • Assemble generated scenes and narration locally with ffmpeg
  • Estimate generation cost before spending credits
  • Edit a faceless video made earlier