ai-infrastructure-litellm
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading official configuration assets from the LiteLLM project repository on GitHub.
- Evidence: Reference to
curl -O https://raw.githubusercontent.com/BerriAI/litellm/main/docker-compose.ymlinreference.md. - [INDIRECT_PROMPT_INJECTION]: The skill configures an infrastructure layer that processes untrusted data which is subsequently passed to LLMs.
- Ingestion points: Data processed via the OpenAI SDK in
usage.tsandexamples/core.md. - Boundary markers: No specific delimiting or 'ignore' instructions are provided in the configuration examples.
- Capability inventory: The proxy facilitates network requests to LLM provider APIs and interacts with a PostgreSQL database for metrics.
- Sanitization: No input validation or content filtering mechanisms are demonstrated in the provided patterns.
Audit Metadata