debug-tools

Pass

Audited by Gen Agent Trust Hub on May 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a systematic debugging workflow without any suspicious code execution or network operations.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill includes defensive guidelines in log-injection.md specifically forbidding the capture of sensitive data such as passwords, tokens, API keys, and PII.
  • [PROMPT_INJECTION]: Instructions are focused on technical debugging processes and do not contain patterns aimed at bypassing agent safety filters or overriding system instructions.
  • [REMOTE_CODE_EXECUTION]: The skill relies on standard development tools (git, grep) and local code analysis. No remote scripts are downloaded or executed.
  • [OBFUSCATION]: Analysis revealed no hidden characters, encoded payloads, or homoglyph-based evasion techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
May 8, 2026, 07:04 PM
Security Audit — agent-trust-hub — debug-tools