mobile-debugging

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides bookmarklets and scripts that fetch debugging libraries from jsDelivr and Unpkg CDNs. These are well-known services for web development resources.
  • [COMMAND_EXECUTION]: Provides shell commands for installing debugging tools (e.g., scrcpy, ios-webkit-debug-proxy) via official system package managers like apt, brew, and scoop.
  • [PROMPT_INJECTION]: The debugging tools described (Eruda, vConsole) inspect DOM, logs, and network traffic from external web pages. This presents a potential surface for indirect prompt injection where a malicious site could influence the agent's behavior during debugging.
  • Ingestion points: SKILL.md (references to external debugging tool integration).
  • Boundary markers: Not present.
  • Capability inventory: Shell execution, Node.js, and Python script automation.
  • Sanitization: Not present.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 01:25 AM