proxy-upgrade-safety
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely composed of markdown documentation and reference materials for security auditing. No executable scripts, subprocess calls, or malicious logic were found across the analyzed files.
- [PROMPT_INJECTION]: The content consists of educational examples and audit checklists. There are no instructions designed to override system prompts, bypass safety filters, or extract sensitive internal agent configuration.
- [DATA_EXFILTRATION]: No hardcoded credentials, sensitive file access patterns, or unauthorized network communication logic were detected. The skill does not interact with the filesystem or external APIs.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts. All references are local to the skill package and provide informational guidance rather than automated execution.
Audit Metadata