chrome-devtools

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation instructs users to install the chrome-devtools-mcp package via npx. This is an official package maintained by the Chrome DevTools team (Google) for browser automation and is a well-known service component.
  • [DYNAMIC_EXECUTION]: The skill includes an evaluate_script tool that allows the execution of arbitrary JavaScript within the browser context. This is a standard capability for DevTools-based automation but grants the agent significant power to interact with page state and authenticated data.
  • [PERSISTENCE]: The skill supports 'Scheduled Tasks' through the Chrome extension, allowing workflows to be executed on a recurring basis (e.g., daily or weekly). This persistence is a documented feature for automating recurring browser interactions.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external web pages, console logs, and network responses, which are untrusted sources.
  • Ingestion points: Browser tabs, page DOM, console output, and network request details (referenced in SKILL.md and references/devtools-mcp.md).
  • Boundary markers: Not explicitly mentioned in the skill instructions.
  • Capability inventory: File system writes (saving extracted data), arbitrary script execution (evaluate_script), and network interaction via the browser.
  • Sanitization: The documentation does not specify sanitization methods for content extracted from web pages before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 11:50 AM
Security Audit — agent-trust-hub — chrome-devtools