development-guidelines

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • Data Exposure (SAFE): The skill interacts with standard project documentation files (docs/*.md). No access to sensitive system paths or credentials was detected.
  • Remote Code Execution (SAFE): No external downloads, package installations, or script execution patterns were found.
  • Prompt Injection (SAFE): The instructions are clear and do not attempt to override safety filters or system prompts.
  • Indirect Prompt Injection (LOW): The skill processes untrusted architectural and repository documents. Evidence: 1. Ingestion points: docs/architecture.md, docs/repository-structure.md. 2. Boundary markers: Absent. 3. Capability inventory: Document generation only. 4. Sanitization: Absent. The risk is considered low as the skill's capability is limited to text-based document generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:19 PM