gooseworks-ai/goose-skills / drafts/marketing/make-video-ads
make-video-ads
VerifiedUse when the marketer says: make me a video ad, a 30-second ad for paid social, a Reel / TikTok / Shorts ad, a UGC-style video made with AI (no real creator), a product demo video, remake this ad I like for my brand, cut our best organic video into an ad, more like my winning video with new hooks, my video ad got rejected — fix it. Delivers a produced, ready-to-post video ad grounded in a researched angle: one message, a hook for the first two seconds, every claim traced to the brand's signed-off claims, no generated person posing as a customer, the exact words and the price approved before anything renders (one video, or one plus two hook variants), an honest review of the result, priced fixes only on a yes, and launch copy and notes for the ads team. Works in chat apps. Never publishes or spends ad budget. Not for real creators filming for you — work-with-creators (unsure? ask: AI-made or a real person); not for a phone-shot organic post — run-organic-social; not for static ads — make-static-ads.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 16.9 KB |
| atoms/present-gate/ATOM.md | markdown | 3.0 KB |
| atoms/review-render/ATOM.md | markdown | 2.4 KB |
| atoms/script-check/ATOM.md | markdown | 2.8 KB |
| atoms/write-script/ATOM.md | markdown | 2.8 KB |
| references/angles-formats-and-company-types.md | markdown | 6.8 KB |
| references/claims-policy-and-people.md | markdown | 7.8 KB |
| references/excellent-script-example.md | markdown | 5.1 KB |
| references/paths-and-gates.md | markdown | 5.7 KB |
| references/reading-the-brain.md | markdown | 6.5 KB |
| references/review-rubric.md | markdown | 5.6 KB |
| references/script-craft.md | markdown | 5.2 KB |
| skill.meta.json | config | 4.3 KB |
| tests/test-by-hand.md | markdown | 8.4 KB |
atoms/write-script/ATOM.md
View on GitHub---
name: write-script
description: Internal step of make-video-ads. Writes 4–5 whole candidate scripts for one committed angle inside the chosen style's plan — same angle type, different hook families, lead first — checks them before saving, saves them, and keeps the checked lead.
---
# write-script
**Reads:** the angle record (Step 3), the style's `plan_limits` and plan note (from `gooseworks:video_create`), the template details (`gooseworks:catalog_fetch` `type: "template"`) when length or ratio matter, the voice-of-customer rows and signed claims (`references/reading-the-brain.md`), brand voice and video learnings, `references/script-craft.md`, `references/excellent-script-example.md`.
**Writes:** the plan via `gooseworks:video_change` (free; `expected_revision` when asked).
1. **Fit the style.** Scenes, speakers, products, word limits, `extra` fields, picture tags, banned phrases, the length range it actually makes, whether it renders persistent on-screen text (a dramatisation label needs it). Set the style's `answers` deliberately — leave tagline and offer empty unless signed or a plain fact. Pick the voice for the audience (`video_formats` `show: "voices"`) when the style takes one. Write inside them.
2. **Mine the sources** for this angle: the pain moment in the customer's words (row ids), the objection, the switching story — or the labelled brand source. These supply texture and the hook, never the benefit.
3. **Hooks:** 4–5 from different hook families; pick the lead by fit with the account.
4. **Lead script** on the beat table for the length; each scene's `line`, `on_screen` (≤ 8 words), `picture` (including which real image is used — a clean packshot), `speaker` (a cast key) where the style has people. Spoken words ≤ ~2.5 per second; product in frame by ~3s; the story reads with the sound off.
5. **Candidates:** 4–5 whole scripts, the lead first, all the committed `angle_type`, each with its own `hook_family` and `brief_field`. Two that say the same thing will be refused (`candidates_too_similar`); if same-type candidates are refused, send the committed angle as the lead and the rest as other angle types, still checked.
6. **Check before saving:** run `atoms/script-check` on every candidate; fix or drop the ones that fail.
7. **Save** with `video_change` (`candidates`, the music mood the style needs, the video `name`). Custom: one timed draft as the server's next step directs.
8. **Keep the checked lead:** read back once as the answer says (the judge takes up to ~90 seconds) and diff the shown words against your checked lead; if it picked another script or rewrote yours, swap your checked lead back (`use_candidate` or whole `scenes`) unless the marketer prefers the judge's — then script-check it and re-set the `name` if the angle or hook family changed. Engine advice that breaks policy is ignored and named at the gate.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |