Local Index

A technical approach that pre-builds a structured index in the user's local environment. By vectorizing codebases or documents through offline preprocessing, it enables AI agents to perform precise, targeted retrieval at inference time, reducing the number of tool calls and token consumption while preserving data privacy.