Cinematic Music Video · عوامی Open Skill
Cinematic Music Video
Cinematic Music Video Digen پر ایک عوامی Open Skill ہے۔ کھولیں اور دیکھیں یہ ورک فلو کیا تخلیق کرتا ہے۔
Music
101 10 42
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/1821775/2026-09-24/gps5gps9.jpg
--- allowed-tools: - transcribe_audio - generate_image - generate_video - mix_clips - create_document - update_document - get_document --- # Cinematic 3D Animated Nasheed Video Turn an uploaded nasheed audio track into a polished full-screen 16:9 cinematic 3D animated video, using the uploaded audio as the **exact** music and timing source (never replace, remix, shorten, or alter it unless asked). Story, characters, gestures, lip-sync, camera, and scenes are all driven by the audio's language, lyrics, section structure, syllables, pauses, and timing. > 🚫 **The religious-movement rule is a hard requirement, never relax it:** the video must contain **no dancing, no choreography, no dance formations, no rhythmic body sway, no hip movement, no dance poses, no stage dancing, no music-video-style dancing**, and characters must **not** synchronise body movement to the beat in a dance-like way. Movement may only be respectful, natural, and purposeful (calm walking, standing, sitting, reading, reflecting, helping, greeting politely, holding a book, opening a door, carrying a lantern, looking toward the sky, hand over the heart, pointing naturally, offering assistance, turning the head, interacting gently with the environment). ## When to use - User provides an uploaded nasheed audio and wants a cinematic 3D animated music/nasheed video from it. - User wants to revise or continue an existing nasheed video package produced by this skill. - User wants a **10-second vertical short-form version** (Reels / Shorts / TikTok) of the nasheed — either instead of or in addition to the full video. ## Reference files - `references/nasheed-video-guide.md` — Detailed creative bible: movement rules, gesture catalogue, facial-expression catalogue, character template, background-world rules, camera-shot catalogue, on-screen-text rules, negative prompt, and visual-quality/consistency checklist. **Read this before the first generation step.** - `references/timeline-template.md` — The time-coded timeline table template (one block per audio section). Read when producing the timed scene plan. ## Steps 1. **Set up a todo list** with `write_todos`: (a) analyse audio, (b) build creative plan + timeline, (c) create characters, (d) create scenes/settings, (e) generate clips with audio-driven lip-sync, (f) stitch final cut (+ optional 10-second short). Update the list as you progress. 2. **Analyse the uploaded audio.** Call `transcribe_audio(audio_url=<uploaded audio asset url>)` (model='asr'). From the transcript and timestamps detect: the **language**, the **intro / verses / chorus / bridge / instrumental sections / pause / ending boundaries**, the **lyrics and syllable timing**, breaths, and vocal emphasis. Record where vocals occur (lip-sync ON) and where they don't (lip-sync OFF — instruments/pauses show breathing/blinking only). Do not invent lyrics; if any lyric is unclear, keep the audio timing without fabricating visible mouth movement. 3. **Build the creative package.** Read the guide, then produce: video title, characters (fill the Character template in the guide — fictional, adult- and child-safe, modest respectful clothing; if the user left fields blank, invent consistent original details), background world (from the user's background description, else a peaceful original environment like a sunrise garden / quiet village / geometric courtyard / mountain / moonlit desert / seaside / family home / school garden / historic street), a **story direction** matching the lyrics (clear beginning → gentle development → meaningful actions → hopeful progression → peaceful ending), camera plan, and on-screen-text plan (minimal; optional opening title, subtitle "An Original Animated Nasheed", closing "Thank you for watching"). Save this as a document via `create_document(name="nasheed-video-plan", entity_type="scripts", content_type="text/markdown")`. 4. **Produce the time-coded timeline.** Using the template in `references/timeline-template.md