platform-builder

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns detected. The skill enforces security best practices for Power Platform development, specifically prohibiting hardcoded credentials and promoting environment-agnostic configuration.
  • [EXTERNAL_DOWNLOADS]: Fetches starter templates from Microsoft's official GitHub organization using npx degit microsoft/PowerAppsCodeApps/templates/starter.
  • [COMMAND_EXECUTION]: Utilizes standard development tools including npm and the Microsoft Power Platform CLI (pac) to scaffold, develop, and build projects.
  • [DATA_EXFILTRATION]: Explicitly includes rules and checklists to prevent the inclusion of secrets, API keys, and hardcoded environment-specific GUIDs in the source code.
  • [PROMPT_INJECTION]:
  • Ingestion points: Processes component specifications and architecture decision records (ADRs) provided in the prompt context (SKILL.md).
  • Boundary markers: Relies on loading specific sub-skills (e.g., code-apps, canvas-apps) to enforce behavioral patterns.
  • Capability inventory: Capable of shell command execution (npm, pac) and writing source code to the filesystem.
  • Sanitization: Includes extensive post-generation self-review checklists to validate output against security and architectural standards.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 02:03 AM
Security Audit — agent-trust-hub — platform-builder