ecommerce-new-arrivals-monitor
Fail
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: Instructions in the 'BrowserAct Runtime' section mandate the automatic installation of the BrowserAct CLI and Skill without user confirmation.
- [EXTERNAL_DOWNLOADS]: The skill requires downloading software from an external site (www.browseract.com) that is not a recognized trusted source.
- [COMMAND_EXECUTION]: The requirement to 'install the missing dependency automatically' implies executing installation scripts or commands silently in the background without user approval.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from arbitrary ecommerce websites which could contain malicious instructions. * Ingestion points: Captures browser-visible page content and metadata (SKILL.md). * Boundary markers: Absent; there are no instructions to isolate or verify external content. * Capability inventory: Network access, workspace file access, and software installation. * Sanitization: Absent; the skill lacks any mechanism to sanitize or filter data from monitored websites.
Recommendations
- AI detected serious security threats
Audit Metadata