what is topic-material-collector fr?

Topic Material Collector is an open source skill for the Codex agent environment that helps someone gather video material for a topic, script, narration, or shot list. You give it a subject, and it searches YouTube first, looks at subtitles and creator chapters to find the exact time ranges that are relevant, then downloads either full videos or just the clipped portions you need. It can also pull in reviewed candidate videos and images from other platforms, rename everything consistently, and write out a manifest of what it collected in Markdown, CSV, and JSON formats. That manifest tracks where each clip came from, who made it, its timestamps, how confident the tool is that it found the right moment, license information, and a label describing how risky it would be to reuse. To run it, you need Python 3.10 or newer, the yt-dlp download tool, ffmpeg and ffprobe for handling video, and either Codex or another agent environment that supports the SKILL.md format the project uses. Installation means cloning the repository and copying its skill folder into your Codex skills directory, after which you can invoke it inside Codex with a command like asking it to collect footage about a specific person or event, or to find supporting clips for a narration script you already have. There are also standalone Python scripts you can run directly outside of Codex to search YouTube, build a catalog of candidates without downloading anything, or download from a manually reviewed list of candidates. The README is explicit that broad searching does not mean covering every page on the internet, and that downloading a video or image does not grant permission to publish it. It also states plainly that the tool does not bypass DRM, paywalls, logins, region restrictions, or watermarks. Collected material gets labeled low, medium, or high risk depending on whether it is clearly public domain or openly licensed, has unclear reuse rights, or looks like protected content such as films or paid news. The project describes itself as a research and organizing tool, not a source of copyright clearance, and tells users to verify rights before using anything publicly or commercially. It is released under the MIT license.