mehbylabs/creator-skills / skills/creator-explainer
creator-explainer
FlaggedTurn a topic, article, notes, or document into a narrated explainer video — script, voiceover, per-scene visuals, music bed, captions, final MP4 — using whatever providers the user has, and degrading gracefully (video clips → animated stills → captioned slideshow). Use when the user asks for an explainer, faceless video, narrated video, YouTube video from an article, educational video, listicle video, or "turn this into a video". Not for single clips (creator-video) or product ads (creator-ad).
Creator Explainer
Project folder: explainer/<slug>/ with script.md, vo/, visuals/, music/,
scenes.json, final.mp4.
Step 1 — Brief (one message, then go)
Infer and state: audience, length (default 60–90 s), format (16:9, or 9:16 for
Shorts/Reels), visual style (one sentence used in every scene prompt), voice. Check routes:
for t in speech image video music; do python3 scripts/creator.py plan --task $t; done
Pick the visual tier from what's available and say it in one line:
| Tier | When | Scene visuals |
|---|---|---|
| A | video provider available | 4–8 s generated clips (first frame from an image when style matters) |
| B | image provider only | generated stills animated with Ken Burns |
| C | neither | text cards (media.py text on a solid/gradient background) |
No speech provider → captions-only video; tell the user which key unlocks narration.
Step 2 — Script (get approval before spending credits)
references/explainer-structure.md: hook (≤ 2 sentences) → 3–6 beats → payoff/CTA.
~150 spoken words per minute. One scene per beat, each with: narration, visual idea,
on-screen text (optional). Save to script.md and show it.
Step 3 — Voiceover
One file per scene: vo/01.mp3 … with creator.py speech; pin provider + voice after the
first success. Scene timing comes from these durations. Non-English: pass --lang; for
Arabic dialects write the script in the dialect and follow references/arabic-voice.md.
Arabic narration: add tashkeel to meaning-changing words (fractions, دَين/دِين، عَقد/عِقد,
passive verbs, numbers in words) before any TTS; speech stops with needs_tashkeel otherwise.
Generate every narration file with --verify and re-voice lines whose numbers or names changed.
Step 4 — Visuals
For each scene, prompt = style sentence + the scene's visual idea.
- Tier A:
creator.py video --prompt … --ratio <r> --duration <≈vo length> --out visuals/NN.mp4(optionally--startfrom an approved still for consistency). - Tier B:
creator.py image --prompt … --ratio <r> --out visuals/NN.png. - Tier C:
ffmpeg -f lavfi -i color=c=#0B1F3A:s=1920x1080 -frames:v 1 bg.pngthenmedia.py text --image bg.png --text "<key line>" --pos center.
Look at each visual before moving on; regenerate off-style ones.
Step 5 — Music (optional) and assembly
creator.py music --instrumental --duration <total+5> --out music/bed.mp3 if available.
[{"visual": "visuals/01.png", "audio": "vo/01.mp3"}, {"visual": "visuals/02.mp4", "audio": "vo/02.mp3"}]
python3 scripts/media.py assemble --scenes scenes.json --out final.mp4 --ratio 16:9 [--music music/bed.mp3]
Captions (recommended for vertical): build subs.srt from the script using scene start
times (cumulative VO durations + 0.4 s), then media.py captions.
Step 6 — Deliver
media.py probe final.mp4; spot-check frames at 3 points. Report length, tier, providers,
and what would upgrade quality (e.g. "add GEMINI_API_KEY for real motion clips").
References
references/explainer-structure.md— script structure and pacingreferences/audio.md,references/video-prompting.md,references/image-prompting.mdreferences/media-ops.md— assemble, captions, text cardsreferences/providers.md
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |