design-orchestration
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues were detected. The skill acts as a meta-instruction set for routing tasks between other tools.
- [PROMPT_INJECTION]: While the skill uses strong imperative language (e.g., "Do NOT allow," "prohibited," "MUST route"), these instructions are intended to define workflow guardrails and state transitions. There are no attempts to bypass system safety filters, perform role-play jailbreaks, or extract system prompts.
- [DATA_EXFILTRATION]: The skill does not access sensitive files, environment variables, or credentials. It lacks any network-capable commands like curl or wget.
- [REMOTE_CODE_EXECUTION]: No external dependencies, package installations, or remote script executions are present. The skill consists entirely of natural language instructions.
- [COMMAND_EXECUTION]: The skill does not invoke shell commands or execute subprocesses. It coordinates other agent skills through logical flow control rather than direct system execution.
Audit Metadata