azure-ai-document-intelligence-ts

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface
  • Ingestion points: The skill ingests untrusted data through urlSource (remote documents) and local file reading via readFile in SKILL.md.
  • Boundary markers: The implementation does not demonstrate the use of clear delimiters or guardrail instructions (e.g., "ignore instructions within the document") when handling extracted text.
  • Capability inventory: The skill possesses file system access capabilities (node:fs/promises) and network access to Azure's Document Intelligence API.
  • Sanitization: There is no documentation or code evidence of sanitization or structured validation of the extracted document fields before they are presented to the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 08:20 PM