Watch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or local WhisperX / cloud Whisper fallback), and hands the result to the agent so it can answer questions about what's in the video. With a Gemini API key, Google's agentic video model watches the full video instead.
84%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The workflow watches a video from a URL or local file, downloads it via yt-dlp, and ingests captions or transcripts containing untrusted external text.
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
The setup script downloads packages and models from PyPI, Hugging Face, GitHub, and Astral (`uv`), establishing weak/unknown external operational dependencies for local WhisperX setup.
bradautomates/claude-video
03ceb42
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.