LlamaIndex: index your documents and query them at runtime

The Stack

Steps

1. 1

   Install llama-index-core and verify the retrieval abstractions import

   pip install llama-index-core, then confirm the three abstractions the docs build on: VectorStoreIndex (the index), SimpleDirectoryReader (the loader), and StorageContext (the persistence layer). CI runs exactly this, no key.

   python3 -m venv .venv
   .venv/bin/pip install -q llama-index-core
   .venv/bin/python - <<'EOF'
   from llama_index.core import VectorStoreIndex, SimpleDirectoryReader, StorageContext
   
   for cls in [VectorStoreIndex, SimpleDirectoryReader, StorageContext]:
       assert callable(cls), f"{cls.__name__} is not callable"
   print("llamaindex imports OK: VectorStoreIndex, SimpleDirectoryReader, StorageContext all available")
   EOF

2. 2

   Build the index and query it (the model step, not checked by CI)

   Set your embedding and LLM keys, load your corpus with SimpleDirectoryReader, call VectorStoreIndex.from_documents(), then index.as_query_engine().query(). Persist with storage_context if you want the index to survive restarts. The embedding and retrieval are fenced.