youtube-serp
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides standard functionality for content and competitor research using the AIsa YouTube API service. No malicious code, obfuscation, or suspicious command execution patterns were found in the provided files.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from YouTube (video titles and descriptions), which is a common attack surface for indirect prompt injection. This is documented as an inherent property of the skill's utility.\n
- Ingestion points: Untrusted metadata is fetched from
https://api.aisa.one/apis/v1/youtube/searchwithin thescripts/youtube_client.pyscript.\n - Boundary markers: The script returns JSON data to the agent context without additional delimiters or specific instructions to ignore instructions embedded in the metadata.\n
- Capability inventory: The skill enables the retrieval of external text data which is then presented to the agent for processing.\n
- Sanitization: The Python client performs standard JSON decoding but does not sanitize or filter the content of the metadata strings before outputting them.
Audit Metadata