debug-mode

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill is a debugging tool that operates entirely within the local environment. It uses a loopback-only logging server (log-server.mjs) to collect data from temporary probes added to the code being debugged.
  • [COMMAND_EXECUTION]: The skill includes scripts that use spawn to manage the lifecycle of its own internal logging server. This is restricted to running the skill's own Node.js scripts in the background.
  • [DATA_EXFILTRATION]: Data collection is strictly limited to 127.0.0.1 (loopback). The documentation in references/probes.md explicitly warns against capturing sensitive information such as secrets, tokens, or personal data.
  • [PROMPT_INJECTION]: No malicious prompt injection or behavior override patterns were detected in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 06:17 AM
Security Audit — agent-trust-hub — debug-mode