youtube-trending-scanner
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves untrusted data from YouTube, including video titles, tags, and channel information, which is then summarized by the agent into a markdown report.
- Ingestion points: The script
scripts/scan_trending.pyfetches data from the YouTube Data API usingsearch.listandvideos.listendpoints. - Boundary markers: The instructions do not define specific delimiters or include warnings to the agent to ignore instructions embedded within the retrieved content.
- Capability inventory: The agent has the capability to execute shell commands (running the script), read JSON data, and write markdown files to the local directory.
- Sanitization: No sanitization or filtering is performed on the YouTube metadata before it is presented to the agent for analysis, allowing for potential manipulation of the agent's output by content creators.
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of an external Python package to interface with the YouTube API.
- Evidence: The instructions specify that the user must run
pip3 install google-api-python-client. This is the official and well-known client library provided by Google.
Audit Metadata