multimodal-llm
Pass
Audited by Gen Agent Trust Hub on May 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a reference for multimodal LLM patterns, providing code examples and architectural guidance for integrating vision, audio, and video APIs. All reviewed content is instructional and uses legitimate API patterns.
- [DATA_EXPOSURE]: Code examples for various APIs (OpenAI, Anthropic, AssemblyAI, Kling) use generic placeholders for credentials (e.g., 'your-api-key', 'YOUR_API_KEY'). No hardcoded secrets or access to sensitive local file paths were found.
- [PROMPT_INJECTION]: The skill describes patterns for processing untrusted multimodal data, such as PDFs and images. It explicitly recommends best practices like using page ranges for large documents to prevent resource exhaustion or context overflow. No instructions for bypassing safety filters or overriding system behavior were identified.
- [EXTERNAL_DOWNLOADS]: The skill references official API endpoints and SDKs for well-known and trusted providers (OpenAI, Anthropic, Google, Kling, fal.ai, AssemblyAI). These references are consistent with the skill's primary purpose and do not involve untrusted remote code execution.
Audit Metadata