browser-qa

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted data from web pages via take_snapshot, wait_for, list_console_messages, and list_network_requests (SKILL.md). No explicit boundary markers or sanitization are mentioned for this content. The skill possesses capabilities such as evaluate_script and bd command execution (SKILL.md). This surface allows for potential indirect prompt injection, which is a common risk factor for browser-based automation tools.
  • [COMMAND_EXECUTION]: The skill executes shell commands using the bd CLI tool for task synchronization and bug reporting. It also includes an evaluate_script capability for running JavaScript within the browser environment for testing purposes. These tools are scoped to the skill's primary function of quality assurance.
  • [SAFE]: No critical vulnerabilities, hardcoded credentials, or unauthorized network activity were detected. The skill's operations are consistent with its stated purpose of browser-based quality assurance.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 03:50 PM