Hook & Caption Pack Skill

What you get

Five opening hooks: slam, counter, question flip, word swap and stack list
Four burned caption styles: word pop, highlight sweep, karaoke fill and chip cut
Two palettes (night volt, paper cobalt) with role overrides and a 4.5:1 text-contrast check
1080×1920 / 30 fps MP4 (15–60 s), cover.jpg, captions.srt, layout.json and render-report.json
Local Kokoro English narration; scene-level editing re-renders only the scene you change

You will need

Node.js 22+, FFmpeg and FFprobe on macOS
HyperFrames 0.8.181 with headless Chrome, and its local Kokoro voice model
A script with a hook line, three to six beats and a closing call to action
Only self-drawn, public-domain, CC0 or properly licensed visual material

Good fit

Tips, lists, recipes and micro-tutorials that need a strong first second
Creators who want to try different hooks and caption looks on the same script
Short scripts of 15–60 seconds with one idea per beat and a closing action
Explainer and how-to channels that want a consistent, poster-like card stage

Not for

Talking-head footage, photo slideshows or screen recordings
Long explainers past 60 seconds or horizontal masters
Unverified news, medical or legal claims
Projects that need bundled music, personal voices or third-party media

You say

“Use Hook & Caption Pack to turn a script into a vertical short with a hook you choose and a caption style you choose. Confirm the hook style, the caption style and the palette, write the beats, then render with local Kokoro narration, burned captions and a scene-level edit handle.”

After installing, say something like this to your agent with your own topic and assets. It asks only for what is missing.

FAQ

Do I need an account?

Yes. Members Skills install from a signed-in FeedSkills account with an active Founding Membership.

What does it cost to run?

The Skill is included with Founding Membership. It needs the tools listed under “You will need”. Any voice, model or media service you choose bills you directly; FeedSkills does not include provider credits.

Can I use the videos commercially?

Videos you make are yours to use, including commercially. The Skill package is licensed as: Source available · original FeedSkills design · fonts SIL OFL 1.1 · upstream technique reference MIT. Fonts, music, footage and services you add keep their own terms.

Which agents does it work with?

Agents that read SKILL.md files, such as Codex or Claude Code. What FeedSkills has actually tested is listed under Tested scope below.

Tested scope

Version: 0.1.3
Acceptance: This version has a recorded acceptance entry.
Tested: Two 23-second 1080×1920 / 30 fps English example films with different combinations (counter hook + karaoke fill captions on the volt palette; word swap hook + highlight sweep captions on the cobalt palette) rendered with HyperFrames 0.8.181 and local Kokoro narration; burned captions plus captions.srt and a DOM-measured layout.json; a one-scene headline edit re-rendered only its target scene and the framediff proof showed the other six scenes frame-identical. The flip and stack hooks and the word-pop and chip-cut captions were checked on short sampler briefs for layout only, not as finished films. User approved both example films on 2026-10-28. This is taste approval of the supplied examples; clean operating-system, physical social-app and real customer adoption checks remain separate. Package 0.1.1 adds atomic project-directory ownership and complete code cache fingerprints; the approved 0.1.0 example video files are retained unchanged. The user explicitly approved reusing the exact approved video files for package 0.1.1 on 2026-10-28; the new code delivery is independently revalidated.
Runs on: Claude Code · Codex · macOS · Node.js 22+ · FFmpeg
License: Source available · original FeedSkills design · fonts SIL OFL 1.1 · upstream technique reference MIT

Details

Two reviewed themes

Hook & Caption Pack is a reusable vertical design with two different English example films. Both were approved by the product owner on October 28, 2026. This confirms review of the examples, not customer adoption or social-platform performance.

Watch Theme A · cover
Watch Theme B · cover

Make your own film

Start from an included example brief. Use a new project directory outside the installed Skill, keep stable scene IDs, write concise English copy and render with local Kokoro narration or an explicitly silent cut. Outputs include MP4, cover, burned captions, SRT, measured layout and a render report. Review the complete film with sound before sharing.

For revisions, name one scene and the change; unchanged scene segments are reused. The local edit example was checked independently with frame comparisons. No paid media generation is required. You need Node.js 22+, FFmpeg/FFprobe and pinned HyperFrames 0.8.181; local narration also needs the Kokoro environment.

Access and boundaries

This is a Members Skill. An active Founding Membership is required; the public catalog contains this summary and the approved previews only. Package files use private, entitlement-checked delivery.

The example films have passed taste review; a fresh operating system, physical social-app overlays and real customer adoption remain separate checks. This package does not bundle personal voices, provider credentials or third-party music.

The finished films exercise counter/karaoke and swap/highlight combinations. The other hook and caption styles have layout tests, but have not all been tested as complete rendered films. Word emphasis is estimated from narration duration, not forced audio alignment.

License and boundaries

Huashu Motion’s kinetic-type technique is a reference at pinned commit 52467f99c602b76ba86d511d37acdfe93fa1e16e; the FeedSkills engine is its own implementation. The Skill preserves upstream licensing and bundled-font notices. Read STYLE.md and THIRD_PARTY_NOTICES.md for the adaptation, permitted materials and limits.