application-modeling

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes administrative commands via the Radius (rad), Kubernetes (kubectl), and Docker CLIs to provision and manage application infrastructure.
  • [EXTERNAL_DOWNLOADS]: Fetches Bicep extensions and infrastructure recipes from well-known OCI registries, including Azure Container Registry and GitHub Container Registry.
  • [EXTERNAL_DOWNLOADS]: Automates the download of the Bicep CLI via the Radius CLI tool.
  • [PROMPT_INJECTION]: The skill processes external configuration files such as the platform constitution and resource manifests. These files define the logic for environment setup and resource provisioning, representing an indirect prompt injection surface. 1. Ingestion points: Platform-Engineering-Constitution.md, manifest.yaml, and Bicep files. 2. Boundary markers: Not present. 3. Capability inventory: Subprocess execution via CLI tools and network operations for artifact retrieval. 4. Sanitization: Not explicitly performed on configuration file contents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 11:41 PM