Turn raw video clips and a logo into a finished branded video for TikTok or Reels with one command.
Generate a researched, fact checked video story about a specific topic using an AI brief.
Review and hand edit the written story before rendering the final video.
Build a branded video from a written config file for full manual control.
| roomcareos/scenecast | 0xallam/my-recipe | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | Python | Python | Python |
| Last pushed | — | 2022-11-22 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 2/5 | 4/5 |
| Audience | general | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires Python 3.11+, ffmpeg installed locally, and a one-time headless browser download for rendering.
Scenecast is a command line tool that turns your own video clips and a company logo into a finished, on-brand short video, without you having to build a timeline or write any code. You give it just two things, a logo file and a folder of clips, and it works out your brand colours from the logo, watches your footage to find the best cuts, writes on-screen text, and renders a finished vertical video ready to post. Everything happens on your own computer. Your media files are never uploaded anywhere, which the project treats as a core design choice rather than an afterthought. Running one command reads your brand from the logo, reviews up to five minutes of each clip, keeps the strongest moments, drops the dull parts, writes the on-screen story text itself, and renders a polished video in your colours with clean transitions between shots. Sound is left out on purpose since you would normally add music or voice later in TikTok or Instagram Reels' own editor. If you want the video to be about something specific, you can give it a short written brief along with a research option, and it will use an AI service with web search to check facts and note where each one came from, using your own Anthropic API key. Even in this mode, only your brief and a short description of your footage are sent out, the actual video files stay local. Every run also saves an editable draft of the story next to the output, so you can review it, give written feedback for a revised version, or hand-edit the wording yourself and re-render it exactly as written. For people who prefer full control from the start, there are separate commands to read your brand into a config file and then build the video from that file directly, plus a guided setup that just asks you a few questions. Scenecast is written in Python and needs Python 3.11 or newer plus ffmpeg installed on your machine, since ffmpeg is the free video engine it uses to read and encode video files. Installation is a pip install, followed by a one time download of a headless browser used for rendering.
A command line tool that turns your logo and video clips into a finished, on-brand vertical video automatically, running entirely on your own machine.
Mainly Python. The stack also includes Python, ffmpeg, Playwright.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.