azure-ai-formrecognizer-java
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- CREDENTIALS_UNSAFE (SAFE): The document explicitly identifies hardcoded credentials as an anti-pattern (Section 2.4) and provides code samples using environment variables and
DefaultAzureCredentialBuilderas the secure alternative. - PROMPT_INJECTION (SAFE): No instructions targeting agent behavior or safety filter bypasses were detected. The content is technical and instructional.
- EXTERNAL_DOWNLOADS (SAFE): While it links to a trusted Azure SDK repository, it does not instruct the agent to download or execute external scripts.
- DATA_EXFILTRATION (SAFE): No network exfiltration patterns or unauthorized sensitive file access commands were found.
Audit Metadata