skills/cauu/skills/codebase-design/Gen Agent Trust Hub

codebase-design

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of markdown documentation and metadata defining a conceptual framework for software design (deep modules, seams, adapters). It contains no executable scripts, external network calls, or commands that interact with the host system.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a process for spawning sub-agents to explore design alternatives based on user-provided codebase context. While this involves processing untrusted data at runtime, the risk is negligible as it is restricted to architectural design synthesis.
  • Ingestion points: DESIGN-IT-TWICE.md (Step 1: User-facing problem space explanation).
  • Boundary markers: Not explicitly defined in the sub-agent prompting logic.
  • Capability inventory: Spawns sub-agents using the platform's native Agent tool to generate design variations.
  • Sanitization: The skill focuses on conceptual design and does not describe specific sanitization of user-provided code descriptions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 03:15 AM
Security Audit — agent-trust-hub — codebase-design