architecture-to-issues

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to read from and write to markdown files within the docs/ai-native/ directory. It does not attempt to access sensitive system paths, environment variables, or user credentials.
  • [COMMAND_EXECUTION]: There is no evidence of subprocess spawning, shell execution, or dynamic code execution. The skill instructions are purely text-based and focused on content generation.
  • [DATA_EXFILTRATION]: The skill lacks network capabilities such as curl, wget, or API requests. All data transformations are performed locally and saved to the filesystem.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from external documentation files.
  • Ingestion points: docs/ai-native/03-prd.md, docs/ai-native/04-architecture.md.
  • Boundary markers: None explicitly defined to delimit the input content.
  • Capability inventory: Limited to reading and writing markdown files within the project scope.
  • Sanitization: No explicit sanitization or validation of the input content is specified.
  • Note: Given the limited capability of the skill (markdown writing only), this surface represents a minimal risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 04:18 AM
Security Audit — agent-trust-hub — architecture-to-issues