You've got a talking head video and want to add designed graphics on top without touching the timeline or audio. This drops titles, lower thirds, data callouts, and quote cards over your clip based on what's being said. The agent writes HTML for each overlay, assembles a composition, and renders through Hyperframes. It's not for subtitles (that's embedded-captions) and won't recolor or reframe your footage. The routing matters here: if you just need spoken words as text, you're in the wrong place. With 29K installs, it's clearly solving a real production gap for people who need polished overlays without opening an NLE.
npx -y skills add heygen-com/hyperframes --skill graphic-overlays --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills