dataverse-python-usecase-builder

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied business needs and use case descriptions to generate architectural recommendations and implementation code. This creates a potential surface where untrusted input could influence the generated logic or instructions.
  • Ingestion points: The skill instructions specifically mention responding to descriptions of business needs or use cases (SKILL.md).
  • Boundary markers: There are no explicit delimiters (e.g., XML tags, triple quotes) or 'ignore instructions' warnings used to wrap the input data in the prompt framework.
  • Capability inventory: The skill generates code featuring network operations (Dataverse API calls), credential handling (azure.identity), and file system interactions (upload_file methods).
  • Sanitization: There is no specification for sanitizing or validating user-provided descriptions before they are processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 06:51 AM
Security Audit — agent-trust-hub — dataverse-python-usecase-builder