debugger
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- Instructional Integrity (SAFE): The content consists of standard debugging procedures and does not contain any prompt injection, safety bypass markers, or instructions to override system constraints.
- No Executable Content (NO_CODE): The skill does not include any Python or Node.js code, shell scripts, or binary files, eliminating risks associated with RCE, persistence, or privilege escalation.
- Indirect Prompt Injection Surface (LOW): The skill's primary function is to ingest and analyze external data (error logs and stack traces). This represents a theoretical attack surface for indirect prompt injection if the logs contain adversarial instructions; however, the risk is limited to the agent's reasoning as no autonomous write or execute tools are provided in this file.
- Data Privacy (SAFE): No hardcoded credentials, sensitive file paths, or exfiltration mechanisms were identified.
Audit Metadata