architect-it
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a comprehensive and professional framework for architectural design, codebase analysis, and documentation. It adheres to established engineering practices and emphasizes clarity, verification, and human oversight for all structural changes.
- [INDIRECT_PROMPT_INJECTION]: As the skill surveys the existing codebase to inform its designs, it is theoretically exposed to indirect prompt injection if malicious instructions are embedded in the analyzed files. This risk is effectively mitigated by the skill's design, which requires producing a human-readable 'Decision' and 'Slice plan' and recording choices in formal documentation (ADRs) before any implementation begins.
- [COMMAND_EXECUTION]: Tool usage is strictly limited to repository exploration, issue management, and the orchestration of read-only subagents for parallel research. The skill's operating contract explicitly requires maintaining harness permission prompts for any risky commands and prohibits autonomous changes to resources outside the repository.
Audit Metadata