architecture-visualization

Pass

Audited by Gen Agent Trust Hub on Sep 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for the agent to install the d2 diagramming tool from well-known sources such as nix shell nixpkgs#d2 or go install oss.terrastruct.com/d2@latest if the command is missing.
  • [COMMAND_EXECUTION]: The skill uses shell commands to carry out its primary purpose, specifically executing the d2 binary to render diagrams, using date for filename generation, and command -v for tool discovery.
  • [INDIRECT_PROMPT_INJECTION]: The skill's functionality involves researching the codebase and interpreting its structure, which represents a surface for indirect prompt injection from untrusted source data.
  • Ingestion points: Codebase research and dependency identification (SKILL.md, Process Step 1).
  • Boundary markers: None identified.
  • Capability inventory: Execution of d2 and date binaries via shell commands, and file writing capabilities within the docs/ directory tree.
  • Sanitization: The skill does not explicitly describe any sanitization of codebase data before it is incorporated into visual diagrams.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 28, 2026, 01:29 AM