finding-youtube-micro-influencers
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides examples of running a local script
scripts/run_actor.jsand usingcurlto interact with the Apify REST API. These are standard development tasks for the skill's stated purpose. - [EXTERNAL_DOWNLOADS]: The skill fetches data from
api.apify.com, which is the official API endpoint for the platform used by the scraper. This aligns with the developer's infrastructure and the skill's purpose. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from YouTube (titles, descriptions, channel names). While this presents a surface for indirect prompt injection where the LLM might process adversarial text from a YouTube video description, the risk is inherent to data scraping and is handled by standard LLM guardrails.
Audit Metadata