ideogram-core-workflow-a
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provide an interface for sending user-defined prompts to an external image generation service, which represents a potential surface for indirect prompt injection if the source data is untrusted. 1. Ingestion points: The prompt argument within the generateImage and generateV3 functions in SKILL.md. 2. Boundary markers: The functions do not implement specific delimiters or safety instructions to prevent the execution of instructions embedded within the prompt text. 3. Capability inventory: The skill has the ability to perform network requests via fetch and write files to the local system using fs.writeFileSync. 4. Sanitization: There is no evidence of prompt sanitization or filtering to mitigate malicious content in the input strings.
Audit Metadata