architect
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: A comprehensive analysis of the skill's instructions, conventions, and design templates revealed no evidence of prompt injection, data exfiltration, or malicious intent. The skill adheres to safe orchestration practices, emphasizing human approval for critical decisions.
- [COMMAND_EXECUTION]: The skill references the 'herdr' orchestration tool for managing subprocesses and delegation. These interactions are governed by defined conventions (e.g., HERDR_ENV triggers) and focus on task management within a development environment, presenting no unexpected security risk.
- [INDIRECT_PROMPT_INJECTION]: The skill's 'Grounding' phase involves ingesting data from the local codebase to inform the design process. While this involves processing external data, the activity is restricted to architectural modeling and is a core requirement of the skill's functional purpose.
Audit Metadata