architect
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and follows standard software architecture practices for generating Architecture Decision Records (ADRs) and design briefs. It does not include any executable code, shell commands, or network operations.
- [DATA_EXPOSURE]: The skill defines a process for writing design documentation to the file system (e.g.,
docs/design/). This is limited to architectural artifacts and does not target sensitive system files or credentials. - [INDIRECT_PROMPT_INJECTION]: While the skill ingests external data (repository history, existing code, and user input) to ground its architectural decisions, this is a standard functional requirement for the task. The process includes 'cross-judging' and human-in-the-loop agreement, which act as safeguards against unexpected behavior derived from processed data.
Audit Metadata