tanstack-ai-core-media-generation
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill integrates with well-known AI services (OpenAI, Google Gemini, BytePlus, fal) and utilizes the established @tanstack ecosystem for media generation.\n- [SAFE]: Documentation includes essential security and logic guidance, such as identifying deprecated functions and correctly handling provider-specific environment variables.\n- [SAFE]: Code patterns for TanStack Start and persistence middleware follow industry standards for secure server-side processing and data lifecycle management.\n- [PROMPT_INJECTION]: The skill design involves processing untrusted user input (prompts, audio, and images) which are forwarded to AI models for processing.\n
- Ingestion points: Prompt and media input fields within SKILL.md.\n
- Boundary markers: The provided examples do not demonstrate the use of input delimiters or explicit instructions to ignore embedded commands.\n
- Capability inventory: The skill utilizes network access to communicate with external AI providers and supports file system writes via optional persistence middleware.\n
- Sanitization: Input data is transmitted verbatim to provider-specific adapters.
Audit Metadata