agent-browser

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install the 'agent-browser' package from the NPM registry and subsequently run an internal installer to download Chromium binaries.
  • [COMMAND_EXECUTION]: All skill functionality is implemented through local command-line execution of the 'agent-browser' utility.
  • [DATA_EXFILTRATION]: The skill includes commands to extract browser state information, such as 'cookies', 'storage local', and 'storage session', which can contain sensitive session data.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from external websites into the agent's context.
  • Ingestion points: Data is ingested via 'agent-browser snapshot', 'get text', and 'get html' commands in SKILL.md.
  • Boundary markers: The skill does not define specific delimiters or instructions to treat web content as untrusted.
  • Capability inventory: The skill can execute CLI commands that interact with the filesystem (PDF/screenshot generation) and the network (browser navigation).
  • Sanitization: No sanitization or filtering of the retrieved web content is specified in the skill's instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 04:47 AM
Security Audit — agent-trust-hub — agent-browser