youtube-feed
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script
scripts/get_updates.pyto retrieve the latest video information. The script allows for user-configurable time ranges via command-line arguments. - [EXTERNAL_DOWNLOADS]: The underlying script performs HTTP requests to YouTube and
r.jina.ai(Jina Reader) to fetch video metadata such as titles, descriptions, and view counts. Jina is a well-known service for processing web content for large language models. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes external, untrusted content from YouTube video metadata.
- Ingestion points: Fetches video titles and descriptions from YouTube and Jina AI in
scripts/get_updates.py. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the prompt instructions.
- Capability inventory: The agent is instructed to translate and summarize the external metadata.
- Sanitization: The script performs basic text cleaning for formatting and length but does not implement security-focused sanitization to prevent instruction injection.
Audit Metadata