web-debug-browser

Warn

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructions and troubleshooting guide recommend running 'npx -y chrome-devtools-mcp@latest', which downloads and executes code from an external registry without author verification or version pinning.
  • [EXTERNAL_DOWNLOADS]: Fetches the 'chrome-devtools-mcp' package from the npm registry at runtime.
  • [COMMAND_EXECUTION]: The 'evaluate_script' tool allows for the execution of arbitrary JavaScript code within the context of the browser.
  • [COMMAND_EXECUTION]: Several tools, including 'take_screenshot', 'take_snapshot', and 'performance_start_trace', accept a 'filePath' parameter, enabling the agent to write files to the local disk.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external websites.
  • Ingestion points: 'new_page' and 'navigate_page' tools (SKILL.md, operations.md)
  • Boundary markers: Absent
  • Capability inventory: 'evaluate_script' (script execution), 'upload_file' (file access), and 'filePath' parameters (file system writes) (tool-reference.md)
  • Sanitization: No sanitization of external web content is described.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 12, 2026, 02:42 PM
Security Audit — agent-trust-hub — web-debug-browser