agent-browser

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted web content through navigation and page analysis commands, which provides an entry point for indirect prompt injection attacks. * Ingestion points: agent-browser open, snapshot, and get text commands in SKILL.md. * Boundary markers: The instructions do not define any delimiters or warning markers to isolate untrusted content from agent instructions. * Capability inventory: The skill has capabilities for element interaction (click, fill, press), file system writes (screenshot, state save), and session management in SKILL.md. * Sanitization: No data validation or content filtering mechanisms are specified for the extracted web information.
  • [COMMAND_EXECUTION]: The skill provides an interface to execute browser automation tasks using the agent-browser command-line tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 02:19 PM
Security Audit — agent-trust-hub — agent-browser