docfactory-arch-expo
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- Role and Instructions (SAFE): The skill is strictly focused on creating a design artifact (04-tech-architecture.md). It explicitly states it does not generate code or migrations, limiting its capability to text generation.
- Data Handling (SAFE): While the skill reads project documents (which might contain sensitive project context like stack decisions), it does not transmit data externally or write to sensitive system paths. It follows the principle of least privilege by requesting only necessary context files.
- Validation Script (SAFE): The included Python script
scripts/validate_docfactory_arch_expo.pyis a simple structural validator that checks for the presence of specific headings and tables in the output file. It uses standard library modules (sys,pathlib) and does not perform network operations or unsafe system calls. - Indirect Prompt Injection (LOW): The skill ingests untrusted data from previous documentation stages (PRD, UI/UX specs). However, since the output is a static markdown file intended for human review and the skill lacks command execution or network capabilities, the risk of an injection causing harm is negligible. No boundary markers are used, which is a minor best-practice omission rather than a vulnerability.
Audit Metadata