clipify-video-clip-generator
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The installation process requires cloning a repository from a personal GitHub account (https://github.com/louisedesadeleer/clipify.git) and installing dependencies including the 'openai-whisper' and 'numpy' Python packages, as well as the 'ffmpeg' system utility.
- [COMMAND_EXECUTION]: The skill performs extensive video and audio processing by executing 'ffmpeg' commands and running local Python scripts (such as analyze.py and build_pan.py) via shell commands and Python's subprocess.run, using paths and parameters derived from user input or processed video data.
- [INDIRECT_PROMPT_INJECTION]: The skill implements logic to scan video transcripts for specific keywords (e.g., 'haha', 'lol', 'crazy') to identify potential video highlights, creating a surface where content in the video transcription could influence the skill's behavior.
- Ingestion points: External video files and the resulting transcription data generated by Whisper (SKILL.md).
- Boundary markers: No explicit delimiters or instructions are used to separate transcription content from the skill's processing logic.
- Capability inventory: The skill can execute shell commands via ffmpeg and run Python scripts that modify local files (SKILL.md).
- Sanitization: Transcription text is converted to lowercase for simple string matching, but no sanitization is applied to the content before it is used to drive the candidate selection logic.
Audit Metadata