browser-use-cloud

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of processing untrusted web content.
  • Ingestion points: User-supplied URLs are passed to the client.run method in SKILL.md, allowing the agent to navigate to and process arbitrary external websites.
  • Boundary markers: The prompt templates do not include explicit delimiters or instructions to the agent to ignore potentially malicious commands embedded in the web pages being browsed.
  • Capability inventory: The skill utilizes the browser-use-sdk which provides extensive capabilities including navigation, form filling, clicking, and data extraction, which could be abused if the agent follows instructions from a malicious site.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the content retrieved from external URLs before it is processed by the AI model.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 10:29 AM