Repository inventory

longstories/movie-maker-skill

Skills indexed from this repository, with install-style signals scoped to the repo.
1 skills0 GitHub stars0 weekly installsGitHubOwner profile

Overview

This skill generates docs-first guidance and artifacts for producing 10–60 second mini-movies locally. It produces numbered scripts, scene-based shotlists, and a render-ready manifest JSON that maps lines to timed shots. The skill focuses on voiceover-first pipelines, provider presets, and practical rendering rules (ffmpeg/Remotion).

How this skill works

Start with a structured intake that captures length, aspect ratio, voice style, character descriptors, music choice, visual style, quality, model preset, and optional reference images. The skill applies rules to convert the script into timed voiceover segments, finalizes shot durations, and outputs a manifest that ties scripts, shots, prompts, and provider hints together. Guidance includes model presets, provider-agnostic prompting, rendering steps, QA checks, and sharing instructions for Clawtube.

When to use it

  • You need a 10–60s mini-movie with script, shotlist, and render-ready manifest.
  • You want a voiceover-first pipeline and clear timing rules for voice and shots.
  • You will render locally with ffmpeg or Remotion and need a JSON manifest to drive the process.
  • You need provider-agnostic prompting and model preset recommendations for cost vs. quality.

Best practices

  • Always run the prescribed intake before generating content to lock choices and references.
  • Write concise numbered script lines (L001…) and map them to scene shotlists (S01…); keep line lengths aligned to target duration.
  • Use the provided model presets: Cheap and Fast for rapid drafts, High Quality for final renders.
  • Generate voiceover timing first, then finalize shot durations to avoid mismatches.
  • Include reference images saved under assets/refs and document any per-shot model overrides in the manifest.

Example use cases

  • Create a 30s product teaser with narrator-only voiceover and cinematic style preset.
  • Produce a 45s 3D animated character vignette with dialogue and character descriptors for consistency.
  • Iterate quickly using Draft quality and Cheap and Fast preset to test framing and pacing before final HQ renders.
  • Prepare a manifest for local rendering and sharing to Clawtube after manual QA and ffmpeg/Remotion assembly.

FAQ

No. It provides rules, scripts, shotlists, and a manifest but does not include pipelines, media storage, or provider keys.

Can I change models per scene or shot?

Yes. The rules include an overrides pattern to specify per-scene or per-shot model presets and quality settings in the manifest.

1 skills

More from this maintainer
Other repositories and skills published under the same GitHub owner.
Skills library
Jump back to the full directory or explore grouped topics.
Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational