planning-decision-engineer
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No patterns of instruction override, safety filter bypass, or system prompt extraction were found. The skill maintains a consistent focus on its technical domain.
- [DATA_EXFILTRATION]: No access to sensitive files or credentials was detected. The skill does not perform any unauthorized network operations or exfiltrate data to external domains.
- [COMMAND_EXECUTION]: There are no instances of arbitrary command execution or shell injection vulnerabilities. Python code blocks are used for algorithmic demonstration and do not involve unsafe system calls.
- [EXTERNAL_DOWNLOADS]: No remote script execution or unverified package downloads were identified. The skill references well-known industry tools (e.g., CARLA, nuPlan, CasADi) for documentation and benchmarking purposes.
- [OBFUSCATION]: No hidden content, encoded payloads, or deceptive characters were found in the skill or its references.
- [DYNAMIC_EXECUTION]: The code samples provided are static examples of planning algorithms and do not involve runtime code generation from untrusted sources.
Audit Metadata