I built a small engine that gives local LLMs real tools

说明与用途

L.A.T.E. 是一个轻量级的 Node.js 引擎，为本地 LLM 提供精确计算、带来源的网络搜索、持久记忆和文件访问功能，降低模型的幻觉和错误率。

问题

Small local LLMs struggle with basic tasks such as math, memory, and web research, often producing incorrect or fabricated answers.

结果

L.A.T.E. enables local LLMs to perform accurate calculations, conduct web research with sources, maintain memory across sessions, and access local files, improving reliability without external paid services.

工作流

• Install Node.js 18+ and run npm start to set up the engine.
• The engine detects local Ollama or LM Studio instances or prompts for an API key for cloud providers.
• The user interacts via chat; the engine routes specific queries (math, web, file, memory, date) to the appropriate tool.
• The model receives tool outputs and crafts the final answer.
• The engine prevents repeated tool calls and limits to five tools per question.
• Custom tools can be registered with ai.registerTool and new providers added in ~15 lines of code.

证据

Small local models are good at chatting and bad at the basics. They get math wrong, they forget everything between chats, and when they don't know something they make it up.

来源

The idea is simple: the tools do the exact work, and the model only writes the answer.

来源

Exact math. The calculator does the maths, not the model. Divide by zero gives a proper error.

来源

Research with sources. It searches, opens the top few pages, picks the part that matches your question, and answers only from what it read, with links.

来源

Memory that survives a restart. Tell it what you're working on, close it, open it again, it still knows.

来源

The first time, it installs everything and looks for a model on your laptop, and asks how you want to connect: ...

来源

Search scrapes DuckDuckGo, so it can get rate-limited if you hammer it.

来源