archiledger

概述

Archiledger 为 AI 助手提供基于知识图谱的内存系统，通过实现模型上下文协议（MCP），使记忆能够跨对话持久化。它既提供低级的手动控制工具，也通过其智能记忆（Agentic Memory）模块提供高级的 AI 驱动内存管理功能。系统使用嵌入式图数据库（LadybugDB），支持向量搜索进行笔记的语义相似性检索，并具有标记、链接和图遍历等功能。

试试问 AI

装完之后，这里有 5 个你可以让 AI 做的事：

• create_notes

  Create one or more memory notes with content, keywords, tags, and optional links

  试试:Create a new note about my project architecture试试:Add a note with links to related requirements
• get_note

  Retrieve a specific note by ID (increments retrieval counter)

  试试:Get the note with ID 12345试试:Show me the note about authentication
• get_notes_by_tag

  Find all notes with a given tag

  试试:Find all notes tagged as 'bug'试试:Show me all notes with the 'architecture' tag
• delete_notes

  Delete notes by their IDs, including associated links and embeddings

  试试:Delete note ID 12345试试:Remove outdated notes about deprecated features
• add_links

  Add typed links between notes with context

  试试:Link note 123 to note 456 with a 'DEPENDS_ON' relationship试试:Connect these two notes as 'RELATED_TO'
• delete_links

  Remove typed links between notes

  试试:Remove the link between note 123 and note 456试试:Delete all CONTRADICTS links from this note
• read_graph

  Read the entire knowledge graph (all notes and links)

  试试:Show me the complete knowledge graph试试:Export all nodes and relationships
• get_linked_notes

  Find all notes directly connected to a given note

  试试:Show all notes connected to ID 123试试:What notes are linked to this requirement?
• get_all_tags

  List all unique tags currently used across notes

  试试:Show me all tags in the knowledge graph试试:What categories are used to organize notes?
• search_notes

  Semantic similarity search with temperature scaling and threshold filtering

  试试:Find notes similar to 'user authentication'试试:Search for content about microservices architecture
• memory_vector_search

  Perform semantic similarity search across memory notes

  试试:Find notes similar to 'payment processing'试试:Search for information about database migrations
• agentic_memory_write

  Store content with automatic AI classification, tagging, and link generation

  试试:Save this note about API design with automatic categorization试试:Store this information with AI-generated links to related concepts

可对比工具

安装

mvn clean package

java -jar mcp/target/archiledger-server-1.0.0-SNAPSHOT.jar

java -Dladybugdb.data-path=./archiledger.lbdb 
     -jar mcp/target/archiledger-server-1.0.0-SNAPSHOT.jar

{
  "mcpServers": {
    "archiledger": {
      "command": "java",
      "args": ["-jar", "mcp/target/archiledger-server-1.0.0-SNAPSHOT.jar"],
      "env": {
        "LADYBUGDB_DATA_PATH": "./archiledger.lbdb"
      }
    }
  }
}

FAQ

Archiledger 适用于生产环境吗？

不适用，README 明确指出此服务器未实现身份验证，使用的是专为本地开发设计的嵌入式图数据库。不推荐在生产环境中使用。

低级和智能记忆 MCP 服务器有什么区别？

低级 MCP 服务器提供直接的手动控制内存操作的工具，而智能记忆 MCP 服务器使用 AI 自动处理笔记的分类、标记和链接，无需手动管理。

archiledger 对比