FilesystemMCP
A local, customizable MCP server for filesystem access that extends the official Anthropic implementation with additional tools and functionality. It enables users to securely read, write, edit, and manage files and directories within specific allowed local paths.
README
FilesystemMCP
A local, customizable MCP (Model Context Protocol) server for filesystem access — forked from the official Anthropic implementation and extended with additional functionality.
Credits
This project is based on the official @modelcontextprotocol/server-filesystem package by Anthropic, licensed under the terms found in the original repository. The original codebase provides a secure, well-structured foundation for filesystem access via MCP.
Why this fork?
The goal of this fork is to extend the original server with additional tools and customizations for personal/local use, while keeping the core security model (allowed directories, path validation) intact.
Setup
1. Install dependencies
cd FilesystemMCP
npm install
2. Build
npm run build
3. Configure Claude
In your claude_desktop_config.json, point the filesystem server to your local build:
{
"mcpServers": {
"filesystem": {
"command": "node",
"args": [
"/path/to/FilesystemMCP/dist/index.js",
"/your/allowed/directory"
]
}
}
}
Restart Claude after saving.
Tools
All tools from the original server are included:
read_text_file/read_file— Read file contents as textread_media_file— Read image or audio files as base64read_multiple_files— Read multiple files in one callwrite_file— Create or overwrite a fileedit_file— Make line-based edits with diff outputcreate_directory— Create directories recursivelylist_directory— List directory contentslist_directory_with_sizes— List with file sizes and sortingdirectory_tree— Recursive JSON tree viewmove_file— Move or rename filessearch_files— Glob-based recursive file searchget_file_info— File metadatalist_allowed_directories— Show configured allowed paths
License
See the original repository for license details.
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。