clean-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands such as
git log,find, and local Python scripts (arch-scan.py) to perform repository analysis and census operations. - [EXTERNAL_DOWNLOADS]: The skill suggests installing and running well-known development utilities, specifically
import-linterviapipxanddependency-cruiservianpx, for automated architectural checks. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes repository data which constitutes an attack surface for indirect instructions.
- Ingestion points: Repository directory structure and file names within
src, as well as commit history metadata retrieved viagit log. - Boundary markers: No specific boundary markers or "ignore" instructions are present in the command execution snippets.
- Capability inventory: File system traversal (
find), file copy (cp), metadata extraction (git log), and shell command execution (python3,npx,pipx). - Sanitization: The skill relies on standard Unix processing tools (
sort,uniq,head) and dedicated architectural linting software.
Audit Metadata