check-architecture-principles

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the gh CLI to interact with the GitHub API for listing and reading architecture principles. This is a standard operation for fetching project-specific documentation from the vendor's repository.
  • [PROMPT_INJECTION]: The skill processes external text from a remote repository. While this presents a surface for indirect prompt injection, the risk is minimal as it targets an official vendor repository.
  • Ingestion points: Markdown files in camunda/product-development/architecture/principles retrieved via gh api or the github MCP server.
  • Boundary markers: None present in the instructions.
  • Capability inventory: Read-only access via gh api. No code execution or write capabilities are defined within this skill.
  • Sanitization: Standard Base64 decoding of API responses is used; no additional sanitization of the fetched documentation content is performed before evaluation.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 02:52 PM
Security Audit — agent-trust-hub — check-architecture-principles