chrome-browser

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The installation guide recommends downloading and installing the @anthropic-ai/chrome-devtools-mcp package. This is a legitimate tool provided by a trusted organization for browser automation.
  • [REMOTE_CODE_EXECUTION]: The skill's core functionality relies on executing arbitrary JavaScript within the browser context via the evaluate_javascript tool. This is the intended primary purpose of the skill for interacting with the DOM and extracting web data.
  • [PROMPT_INJECTION]: The skill processes untrusted content from external websites via tools like get_page_content and evaluate_javascript, creating a surface for indirect prompt injection where instructions hidden in web content could attempt to influence agent behavior.
  • Ingestion points: get_page_content and evaluate_javascript in references/browser-api.md.
  • Boundary markers: The instructions do not specify the use of delimiters or boundary markers to isolate ingested web content.
  • Capability inventory: The skill has the ability to navigate to arbitrary URLs, execute scripts, and capture screenshots.
  • Sanitization: There is no evidence of sanitization, filtering, or validation of the content retrieved from websites before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 04:05 PM