ln-640-pattern-evolution-auditor
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill coordinates a multi-phase audit of architectural patterns using local file analysis and research tools in a manner consistent with professional development.
- [COMMAND_EXECUTION]: To verify implementation integrity, the skill uses shell commands like npm run build and python -m py_compile on the project being audited to confirm pattern presence.
- [EXTERNAL_DOWNLOADS]: The skill performs external queries via WebSearch and MCP Research to retrieve architectural guidelines and open-source alternatives. No binary downloads or script piping were detected.
- [PROMPT_INJECTION]: The skill ingests external project documentation such as ADRs and Guides for context. While this represents a surface for indirect injection, the risk is mitigated by the skill's use of this data for informational scoring rather than autonomous logic control.
Audit Metadata