Mnehmos Synch
Provides persistent context synchronization and memory management for AI agents across sessions and projects, including file indexing, bug tracking, spatial navigation, and agent-to-agent handoff coordination.
README
mnehmos.synch.mcp
Global Memory Bank for AI Agents
An MCP server that provides persistent context synchronization for AI agents across sessions and projects.
Features
- Active Context - Get/set current working state per project
- Filing Cabinet - Index files with summaries and metadata for fast retrieval
- Memory Search - Search across all indexed content
- Spatial Map - "PC as Rooms" folder navigation metaphor
- Bug Tracking - Log and resolve bugs for agent workflows
- Lock Management - Concurrent agent access coordination
- Context Events - Agent-to-agent handoff protocol
Installation
git clone https://github.com/Mnehmos/mnehmos.synch.mcp.git
cd mnehmos.synch.mcp
npm install
npm run build
Configuration
Add to your MCP client config:
{
"mcpServers": {
"mnehmos.synch.mcp": {
"command": "node",
"args": ["F:\\Github\\mnehmos.synch.mcp\\dist\\index.js"]
}
}
}
Tools
| Tool | Description |
|---|---|
get_active_context |
Get current summary/focus for a project |
set_active_context |
Update active context state |
file_to_cabinet |
Index a file with summary and metadata |
get_from_cabinet |
Retrieve indexed file info |
search_memory |
Search across all indexed content |
list_projects |
List all projects in memory bank |
get_spatial_map |
Get folder structure as "rooms" |
add_room |
Add folder to spatial map |
link_rooms |
Connect two folders |
log_bug |
Log a bug for later fixing |
get_bugs |
Get bugs by project/status |
resolve_bug |
Mark bug as resolved |
acquire_lock |
Lock a resource for exclusive access |
release_lock |
Release a held lock |
get_lock_status |
Check lock state |
emit_context_event |
Emit handoff/checkpoint/error events |
get_context_events |
Get recent context events |
Usage
// Set active context for a project
await client.callTool("set_active_context", {
project_id: "my-app",
summary: "Working on authentication module",
focus: "src/auth/login.ts"
});
// Index a file
await client.callTool("file_to_cabinet", {
project_id: "my-app",
file_path: "src/auth/login.ts",
summary: "Login handler with JWT validation",
key_exports: ["login", "validateToken"]
});
// Search memory
await client.callTool("search_memory", {
query: "authentication",
project_id: "my-app"
});
// Emit handoff event
await client.callTool("emit_context_event", {
project_id: "my-app",
agent_id: "agent-1",
event_type: "handoff",
summary: "Completed auth module, ready for testing"
});
Data Storage
Data is stored in SQLite at:
- Windows:
%APPDATA%\mnehmos-synch\ - macOS/Linux:
~/.config/mnehmos-synch/
License
MIT
Author
Mnehmos
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。