skills/ben2pc/auriga-cli/arch-design/Gen Agent Trust Hub

arch-design

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes git rev-parse --show-toplevel to programmatically determine the repository root. This allows the agent to accurately locate and adhere to project-specific architectural rules and constraints.
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the system's open command to display a generated arch-overview.html file. This is a local-only operation intended to provide the user with a visual representation of the proposed architecture.
  • [PROMPT_INJECTION]: The skill is designed to ingest and process external data, including user-defined specifications and existing source code. This represents a potential surface for indirect prompt injection. However, the skill mitigates this risk through a rigorous, multi-stage workflow that requires candidate comparison, trade-off analysis, and explicit human-in-the-loop selection of the final design direction.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 03:02 AM
Security Audit — agent-trust-hub — arch-design