agent-browser

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill is well-documented with a strong focus on security best practices for browser automation.- [COMMAND_EXECUTION]: The skill uses the agent-browser CLI to perform web interactions. It provides secure patterns for JavaScript execution, such as using Base64 encoding or stdin to prevent shell injection issues.- [CREDENTIALS_UNSAFE]: While the skill handles authentication, it provides clear guidance on using environment variables, an encrypted auth vault, and excluding state files from version control to prevent credential leakage.- [PROMPT_INJECTION]: The skill identifies the surface for indirect prompt injection via untrusted web pages (Ingestion: Browser snapshot output from target URL in SKILL.md; Boundary markers: AGENT_BROWSER_CONTENT_BOUNDARIES; Capability inventory: Full browser control via agent-browser CLI; Sanitization: Domain allowlisting and content marking).
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 01:13 PM
Security Audit — agent-trust-hub — agent-browser