NYC

agent-browser

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill's primary function is to capture and process data from external URLs, creating a surface for indirect prompt injection. * Ingestion points: Website content is captured via agent-browser get text, get title, and snapshot in templates/capture-workflow.sh. * Boundary markers: Absent. The templates do not include delimiters or instructions to isolate scraped content from the agent's control logic. * Capability inventory: The skill has broad browser interaction capabilities including click, fill, and state management in templates/authenticated-session.sh and form-automation.sh. * Sanitization: Absent. No evidence of content filtering or validation before the agent handles the scraped data.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:37 PM