Basic Memory：Basic Machines 推出的创新知识管理系统

Basic Memory: An Innovative Knowledge Management System Launched by Basic Machines

Basic Memory 是由 Basic Machines 推出的一款知识管理系统，它借助与 Claude 等大语言模型（LLM）的自然对话来构建持久的知识，并将这些知识存储在本地的 Markdown 文件中。通过 MCP 协议，它实现了大语言模型对本地知识库的读写操作。

Basic Memory is a knowledge management system launched by Basic Machines. It leverages natural conversations with large language models (LLMs) such as douyinkameng.com Claude to build persistent knowledge, and stores this knowledge in local Markdown files. Through the MCP protocol, it enables LLMs to read from and write to the local knowledge base.

展开剩余81%

现有大语言模型互动的局限与 Basic Memory 的优势

Limitations of Existing LLM Interactions and the Advantages of Basic Memory

大多数大语言模型的互动是短暂的，并且缺乏背景知识，现有的解决方法存在一定的局限性。而 Basic Memory 具有显著的优势：知识存储在本地，由用户掌控；支持用户与大语言模型双向读写同一个文件；采用 Markdown 和语义模式，结构清晰简单；能够形成可遍历的知识图谱，便于大语言模型跟踪链接；兼容 Obsidian 等编辑器；其基础设施轻量，仅对本地文件进行索引。

Most interactions with large language models are transient and lack background knowledge, and existing solutions have certain limitations. However, Basic Memory has significant advantages: Knowledge is stored locally and controlled by the user; it supports two-way reading and writing of the same file by both the user and the LLM; it adopts Markdown and semantic dy.douyinkameng.com modes, with a clear and simple structure; it can form a traversable knowledge graph, making it easy for the LLM to track links; it is compatible with editors such as Obsidian; and its infrastructure is lightweight, only indexing local files.

实际应用案例：探索咖啡冲泡方法

Practical Application Case: Exploring Coffee Brewing Methods

在实际应用中，以探索咖啡冲泡方法为例，用户可以先与大语言模型聊天分享自己的心得，然后让大语言模型构建知识，生成包含语义和链接的 Markdown 文件，用户可以实时查看并同步这些文件。在聊天时参考相关话题，大语言模型能够构建丰富的上下文，实现人与大语言模型对文件的双向操作，并且知识保留在本地。

In practical applications, taking the exploration of coffee brewing methods as an example, users can first chat with the LLM and share their experiences, and then ask the LLM to build knowledge and generate Markdown files containing semantics and links. Users can view douyinyewu.com and synchronize these files in real time. By referring to relevant topics during the chat, the LLM can build a rich context, enabling two-way operations on the files between humans and the LLM, and the knowledge remains local.

技术实现细节

Technical Implementation Details

从技术角度来看，Basic Memory 将内容存储在 Markdown 文件中，并使用 SQLite 数据库进行搜索索引。它从 Markdown 文件中提取语义，形成 Entity 对象，该对象包含 Observations（观察）和 Relations（关系），以构建知识图谱。它提供双向同步功能，实现了 MCP 协议，并且可以跨工具引用实体。文件格式包括前言、Observations 和 Relations。

Technically, Basic Memory stores content in Markdown files and uses an SQLite database for search indexing. It extracts semantics from Markdown files to form Entity objects, which contain Observations and Relations to build a knowledge graph. It provides two-way dy.douyinyewu.com synchronization, implements the MCP protocol, and allows for the cross-tool referencing of entities. The file format includes a preface, Observations, and Relations.

与 Claude Desktop 的配合及开源特性

Collaboration with Claude Desktop and Open Source Features

该系统与 Claude Desktop 配合使用，在配置好之后可以进行知识同步。大语言模型能够使用多种工具，用户可以尝试不同的提示指令。Basic Memory 是开源的，采用 AGPL -3.0 许可证，为知识管理带来了一种新的方式，助力用户打造个性化的知识库。

This system works in conjunction with Claude Desktop, and after proper configuration, knowledge synchronization can be carried out. The LLM can xhs.douyinyewu.com use a variety of tools, and users can try different prompt instructions. Basic Memory is open source and licensed under AGPL -3.0, bringing a new approach to knowledge management and helping users build personalized knowledge bases.

发布于：浙江省