youtube-video-api-skill

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill takes arbitrary YouTube channel URLs from users and posts them to the BrowserAct workflow API (see SKILL.md input parameters and scripts/youtube_video_api.py which calls https://api.browseract.com/v2/workflow and returns/parses task_info['output']['string']), causing the agent to ingest untrusted, user-generated YouTube content that the agent is expected to read and act on.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The script makes runtime calls to https://api.browseract.com/v2/workflow (e.g., run-task-by-template with TEMPLATE_ID "82163578680973165") which triggers execution of a remote workflow/template and the skill requires a BROWSERACT_API_KEY to operate, so the external URL executes remote code relied upon by the agent.