navigation-architect
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- Prompt Injection (SAFE): No patterns found targeting the override of agent behavior or safety guardrails.
- Data Exfiltration (SAFE): No access to sensitive file paths (~/.ssh, ~/.aws) or hardcoded credentials detected.
- External Downloads (SAFE): No remote scripts or packages are downloaded or referenced.
- Command Execution (SAFE): The skill does not perform any shell commands or system-level operations.
- Indirect Prompt Injection (SAFE): The skill lacks an execution surface (like subprocess.run or eval) that could be exploited by processing untrusted data.
Audit Metadata