enhance-arch-boundaries
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes repository folder structures, file contents, and outputs from other analysis tools (e.g.,
audit-backend-architecture) to determine layering and import rules. This ingestion of potentially untrusted data from the repository creates a surface for indirect prompt injection. - Ingestion points: Local file system (folder structure, source code) and external skill outputs.
- Boundary markers: None specified for the data being analyzed.
- Capability inventory: The skill executes shell commands to install and run architectural linters, and writes to configuration files, documentation (ADRs), and agent rule files.
- Sanitization: No specific sanitization or escaping of the ingested code/structure data is mentioned before interpolation into instructions or configuration files.
- [COMMAND_EXECUTION]: The skill instructions involve executing shell commands to install and configure architectural enforcement tools such as
dependency-cruiser,eslint-plugin-boundaries, andimport-linter. While these are standard development tools, the skill requires the ability to modify CI configurations and local hooks.
Audit Metadata