ideogram-sdk-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements architectural patterns for a secure Ideogram REST adapter, prioritizing the redaction of sensitive credentials, prompts, and binary data from logging and error handling.
- [SAFE]: External documentation references point to official, well-known domains (ideogram.ai and developer.ideogram.ai), which is considered a safe practice.
- [SAFE]: The instructions incorporate defensive programming practices, including schema pinning, strict type definitions to prevent invalid field combinations, and bounded retry logic.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing external documentation to generate code using Write and Edit tools. This ingestion point is considered safe due to the use of official service domains, and the risk is mitigated by explicit instructions for output sanitization, validation with tests, and human review.
Audit Metadata