llamaindex
Pass
Audited by Gen Agent Trust Hub on Mar 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides documentation and code examples for the LlamaIndex framework. It includes standard practices for data ingestion, vector storage, and agent creation using trusted libraries such as llama-index, openai, and anthropic.
- [EXTERNAL_DOWNLOADS]: The documentation references official repositories and services including GitHub, LlamaHub, and OpenAI for legitimate data retrieval and model integration. All external resources are consistent with the skill's stated purpose of building RAG applications.
- [PROMPT_INJECTION]: The skill implements retrieval-augmented generation (RAG) which involves processing external data from sources like GitHub, the web, and local directories. While this establishes a surface for indirect prompt injection, the provided examples use standard boundary markers in PromptTemplate objects to separate context from user queries. The capability inventory includes tool execution via FunctionAgent and LLM invocation via OpenAI and Anthropic integrations.
Audit Metadata