Lore FAQs

Question 1

What technologies does Lore use for its search capabilities?

Accepted Answer

Lore employs a local hybrid semantic and full-text search engine. It's a lightweight Rust binary with a built-in SQLite database, utilizing FTS5 for lexical search and sqlite-vec for vector similarity. It relies on Ollama for generating embeddings locally.

Question 2

What kind of knowledge can I store in Lore?

Accepted Answer

You can store any engineering patterns, conventions, code snippets, architectural decisions, or documentation as markdown files within a Git repository. Lore indexes this content with hybrid full-text and vector search, making it highly discoverable.

Question 3

What is Lore and how does it help developers?

Accepted Answer

Lore is a local semantic search engine for your engineering patterns and conventions. It helps developers by providing contextual wisdom to AI coding agents like Claude Code, ensuring they consult your organization's best practices before writing code. Your knowledge base is managed as simple markdown files in a Git repository.

Question 4

How does Lore integrate with AI coding agents?

Accepted Answer

Lore integrates as an MCP (Multi-Modal Command Protocol) tool, specifically designed for Claude Code and other MCP-compatible clients. This allows AI agents to seamlessly query your internal knowledge base for relevant patterns and conventions, providing them with essential context.

Question 5

Can I use Lore without an AI coding agent?

Accepted Answer

Yes, Lore includes a comprehensive CLI tool. You can use it to initialize your knowledge base, ingest markdown files, and perform direct searches from the command line, allowing you to access your engineering wisdom independently.

Lore

Lore

主要功能

使用案例

主要功能

使用案例