dev-testing
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The utility script
scripts/with_server.pyusessubprocess.Popenwithshell=Trueto execute server commands passed via the--serverargument. This allows for arbitrary command execution within the agent context. - [PROMPT_INJECTION]: Browser automation examples in the
examples/directory ingest untrusted data from external web pages (DOM text, attributes, and console logs) without sanitization or boundary markers. This exposes the agent to indirect prompt injection risks from malicious web content. The agent's capability to execute subprocesses and write files increases the potential impact. - [DATA_EXFILTRATION]: Example scripts like
examples/console_logging.pyandexamples/static_html_automation.pywrite artifacts such as screenshots and console logs to/mnt/user-data/outputs/and/tmp/. While used for legitimate testing artifacts, these locations can be used to store or expose sensitive data processed during automation tasks.
Audit Metadata