Skills
skills/stevengonsalvez/agents-in-a-box/debug-bridge/Gen Agent Trust Hub

debug-bridge

Warn

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill explicitly includes eval and js capabilities, enabling the agent to run arbitrary code within the browser context of the connected web application.
  • [DATA_EXFILTRATION]: The bridge captures and transmits sensitive application data, including full network request/response bodies, console logs, cookies, and localStorage contents.
  • [COMMAND_EXECUTION]: The runbook provides instructions for executing shell commands to manage a local WebSocket server using tmux and npx, including killing processes on specific ports.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the debug-bridge-browser SDK and the use of the debug-bridge-cli via npx from the public registry.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it ingests untrusted data from the web application's DOM and network traffic (e.g., via the ui and network capabilities) without explicit boundary markers or sanitization in the instructions, which could be used to influence the agent's behavior.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 5, 2026, 09:05 AM