Claude Server MCP
克劳德服务器 MCP
⚠️重要提示:项目状态⚠️
该项目处于早期开发阶段(v0.1.0),尚未准备好投入生产使用。目前正在进行重大重写,以解决几个关键问题。请查看“问题”页面,了解当前的限制和计划的改进。
我们建议等待稳定版本(v0.2.0+)发布后再在任何关键工作流程中使用它。
模型上下文协议 (MCP) 服务器为 Claude 提供复杂的上下文管理功能,支持跨会话的持久上下文、特定于项目的上下文组织和对话连续性。
当前限制
服务器目前与 Claude Desktop 以外的 MCP 客户端存在兼容性问题
如果没有特定的项目 ID,上下文列表功能将受到限制
安全功能很少,尚未投入生产
错误处理是基本的,可能无法提供有用的指导
尚未建立测试基础设施
Related MCP server: Threadminder MCP
发展路线图
该项目正在积极改进中。即将推出的主要增强功能包括:
稳定性改进——修复主目录解析和上下文列表的核心问题
增强的错误处理——更好的错误消息和恢复机制
安全增强——输入验证、路径清理和数据保护
高级上下文管理——版本控制、搜索和更好的组织
如需更详细的路线图,请参阅我们的综合分析分支。
特征
项目上下文管理
层次化上下文组织
亲子关系
上下文之间的交叉引用
项目特定的元数据
对话连续性
基于会话的上下文跟踪
对话链
元数据丰富的上下文存储
灵活的标记系统
高效存储
有组织的目录结构
基于 JSON 的存储
快速查找索引
异步操作
安装
服务器会自动在您的 Claude 桌面应用的 MCP 设置中配置。所有上下文都存储在~/.claude/中,以便更好地组织:
~/.claude/
├── contexts/ # General conversation contexts
├── projects/ # Project-specific contexts
└── context-index.json # Quick lookup index工具
项目上下文管理
// Save project context
use_mcp_tool({
server_name: "claude-server",
tool_name: "save_project_context",
arguments: {
id: "feature-design-v1",
projectId: "my-project",
content: "Design discussion...",
parentContextId: "requirements-v1",
references: ["api-spec-v1"],
tags: ["design"],
metadata: { status: "in-progress" }
}
});对话管理
// Save conversation context
use_mcp_tool({
server_name: "claude-server",
tool_name: "save_conversation_context",
arguments: {
id: "chat-2024-01-01",
sessionId: "session-123",
content: "Discussion content...",
continuationOf: "previous-chat-id",
tags: ["meeting"]
}
});上下文检索
// Get context
use_mcp_tool({
server_name: "claude-server",
tool_name: "get_context",
arguments: {
id: "feature-design-v1",
projectId: "my-project"
}
});
// List contexts
use_mcp_tool({
server_name: "claude-server",
tool_name: "list_contexts",
arguments: {
projectId: "my-project",
tag: "design",
type: "project"
}
});文档
上下文管理指南- 上下文类型和使用的详细指南
架构概述- 技术实现细节
使用指南- 一般使用说明
Claude Desktop Integration - 与 Claude Desktop 集成
发展
克隆存储库
安装依赖项:
npm install构建服务器:
npm run build服务器将构建到
build/index.js
配置
服务器通过 Claude 桌面应用程序的配置文件进行配置: ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"claude-server": {
"command": "node",
"args": ["/path/to/claude-server/build/index.js"]
}
}
}贡献
欢迎贡献!请随时提交问题和拉取请求。
执照
麻省理工学院
Available Tools
4 toolsget_contextC
Retrieve context by ID and optional project ID
Inputs: id (required), projectId.
list_contextsC
List contexts with filtering options
Inputs: projectId, tag, type.
save_conversation_contextC
Save conversation context with continuation support
Inputs: id (required), sessionId (required), content (required), continuationOf, +2 more.
save_project_contextC
Save project-specific context with relationships
Inputs: id (required), projectId (required), content (required), parentContextId, +3 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
Each tool has a clearly distinct purpose with no overlap: get_context retrieves a specific context, list_contexts lists contexts with filters, save_conversation_context saves conversation-specific context, and save_project_context saves project-specific context. The descriptions clearly differentiate between retrieval, listing, and two types of saving operations.
All tools follow a consistent verb_noun pattern (get_context, list_contexts, save_conversation_context, save_project_context) with clear, descriptive names. The naming convention is uniform throughout the set, making it easy to understand each tool's function at a glance.
With 4 tools, the count is reasonable for a context management server, covering key operations like retrieval, listing, and saving. It feels slightly minimal but well-scoped, as each tool serves a distinct and necessary function without redundancy.
The tool set provides good coverage for context management with get, list, and save operations for both conversation and project contexts. A minor gap exists in update or delete functionality for contexts, but agents can likely work around this by re-saving or managing contexts through the provided tools.
Maintenance
Related MCP Connectors
Persistent context for Claude. Your AI always knows your projects and next actions across sessions.
Share one project context across ChatGPT, Claude, Telegram and any MCP client.
Persistent memory for Claude Code and Cursor. Stop re-explaining your project every session.
Cross-LLM persistent memory: store context once, recall it from any AI model.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceProvides persistent context management for Claude AI coding assistants, allowing you to save and restore conversation context, create checkpoints, and organize information across sessions to prevent losing important work history and decisions during long coding sessions.243 npm136MIT

Threadminder MCPofficial
FlicenseNot gradedqualityDmaintenanceA persistent context layer for Claude that automatically remembers project status, decisions, and next actions across sessions, eliminating the need to re-explain context.-- AlicenseNot gradedqualityBmaintenanceEnables persistent context management for Claude AI coding assistants by saving, searching, checkpointing, restoring, and sharing project context across sessions and parallel conversations.243 npmMIT
- AlicenseNot gradedqualityAmaintenanceProvides persistent context management for Claude AI coding assistants, enabling users to save, search, checkpoint, restore, and share project context across sessions and parallel sessions. It preserves work history, decisions, and progress so coding workflows survive compaction and context limits.243 npmMIT