Which directories does the server have access to?
Only the directories specified via command-line arguments at startup. The server will refuse operations outside these paths.
@ysthink
基本信息
The SSE version of the MCP service is modified from the Filesystem MCP server
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"Filesystem-MCP-Server-SSE": {
"command": "docker",
"args": [
"build",
"-t",
"mcp/filesystem",
"-f",
"Dockerfile",
"."
]
}
}
}Read complete contents of a file
Read multiple files simultaneously
Create new file or overwrite existing (exercise caution with this)
Make selective edits using advanced pattern matching and formatting
Create new directory or ensure it exists
List directory contents with [FILE] or [DIR] prefixes
Move or rename files and directories
Recursively search for files/directories
Get detailed file/directory metadata
List all directories the server is allowed to access
Filesystem MCP Server SSE is a Node.js server that implements the Model Context Protocol (MCP) for filesystem operations, allowing AI assistants to read, write, and manage files and directories. It is intended for users who need sandboxed filesystem access via MCP.
Install dependencies with npm install, build, and run the server with the allowed directories as arguments (e.g., node .\dist\index.js C:\Users\User\Desktop\project). Alternatively, use Docker by building the image and mounting directories to /projects. Configure your MCP client (e.g., Cherry Stdio Desktop) to point baseUrl to the server's SSE endpoint.
Only the directories specified via command-line arguments at startup. The server will refuse operations outside these paths.
Use the `edit_file` tool with the `dryRun` parameter set to `true`. The server returns a diff and match information without modifying the file.
The `move_file` operation fails if the destination already exists, to prevent accidental overwrites.
Build with `docker build -t mcp/filesystem .` and run with `docker run -i --mount type=bind,src=...` mounting allowed directories under `/projects`. See the README for full command.
Yes, use the `read_multiple_files` tool with an array of paths. Failed reads do not stop the entire operation.

DeviceLink lets you share folders from your computer with AI assistants, securely and in real-time. Nothing is uploaded. Your files stay on your machine, and are only accessible while the device sharing them is connected
Markdown To PDF Conversion MCP
📄 The PDF intelligence layer for AI agents — Agent Document Twin, evidence-first extraction, visual crops, OCR provenance, trust reports, and benchmark-gated releases. MCP server for Claude, Cursor, VS Code, and any MCP client.
MCP server to integrate AWS S3 and LLM
评论