FieldCure MCP Filesystem Server
A secure MCP server that exposes local filesystem operations to AI clients with sandboxed access and runtime directory changes.
README
FieldCure MCP Filesystem Server
A secure Model Context Protocol (MCP) server that exposes local filesystem operations to AI clients. Built with C# and the official MCP C# SDK.
Features
- 17 filesystem tools: read, write, append, delete, search, copy, move, convert-to-markdown, and more
- Sandboxed access: all operations restricted to explicitly allowed directories
- MCP roots protocol: clients can change allowed directories at runtime without restarting the server
- Security hardened: path traversal prevention, symlink resolution, NTFS ADS blocking, Windows reserved name rejection
- Atomic writes: temp-file-then-rename pattern prevents data corruption
- Binary detection: automatically returns base64 for binary files, UTF-8 for text
- Built-in document parsing: supported document formats can be read directly and converted to markdown on disk
- Stdio transport: standard MCP subprocess model via JSON-RPC over stdin/stdout
Installation
dotnet tool
dotnet tool install -g FieldCure.Mcp.Filesystem
From source
git clone https://github.com/fieldcure/fieldcure-mcp-filesystem.git
cd fieldcure-mcp-filesystem
dotnet build
Requirements
- .NET 8.0 Runtime or later
Configuration
Claude Desktop
{
"mcpServers": {
"filesystem": {
"command": "fieldcure-mcp-filesystem",
"args": [
"C:\\Users\\me\\Documents",
"C:\\Projects"
]
}
}
}
VS Code
{
"servers": {
"filesystem": {
"command": "fieldcure-mcp-filesystem",
"args": [
"${workspaceFolder}"
]
}
}
}
Tools
File Operations
| Tool | Description |
|---|---|
read_file |
Read file contents (text as UTF-8, binary as base64) |
read_multiple_files |
Read multiple files at once with per-file error handling |
read_file_lines |
Read a specific range of lines |
convert_to_markdown |
Convert one supported document file into a markdown file on disk |
convert_directory_to_markdown |
Batch-convert supported document files in a directory to markdown files |
write_file |
Create or overwrite a file with atomic write |
append_file |
Append content to end of file with auto-newline |
modify_file |
Find and replace text (plain text or regex) |
copy_file |
Copy a file or directory recursively |
move_file |
Move or rename a file or directory |
delete_file |
Delete a file or directory |
Directory Operations
| Tool | Description |
|---|---|
list_directory |
List contents with type markers, sizes, and dates |
create_directory |
Create a directory recursively |
directory_tree |
Hierarchical tree view with configurable depth |
Search & Info
| Tool | Description |
|---|---|
search_files |
Find files by glob pattern |
search_within_files |
Search text content across files with line numbers |
get_file_info |
File or directory metadata |
MCP Roots Protocol
The server supports the MCP roots protocol for runtime directory changes. When the client changes roots:
- The server requests the updated roots list.
- Allowed directories are replaced entirely.
- CLI args provide the initial sandbox.
- An empty roots list means all file access is denied.
Security Model
All paths are validated through IPathValidator before any filesystem operation:
- Allowed directories define the sandbox boundary.
.and..are normalized viaPath.GetFullPath().- Symlinks must resolve within allowed directories.
- Prefix matching is directory-separator-aware.
- NTFS alternate data streams are blocked.
- Windows reserved names are rejected.
- Allowed directory updates are thread-safe.
The markdown conversion tools validate both source and output paths through the same sandbox rules. They write converted markdown directly to disk instead of returning full document text through MCP, which is usually more token-efficient than read_file plus write_file.
Project Structure
src/FieldCure.Mcp.Filesystem/
|- Program.cs
|- Security/
| |- IPathValidator.cs
| \- PathValidator.cs
|- Tools/
| |- FileOperationTools.cs
| |- DirectoryTools.cs
| \- SearchTools.cs
\- Utilities/
|- EncodingDetector.cs
\- FileSize.cs
Development
dotnet build
dotnet test
dotnet pack src/FieldCure.Mcp.Filesystem -c Release
Limitations
| Supported | Not Yet Supported |
|---|---|
| Text file read/write (UTF-8) | Non-UTF-8 encoding auto-detection |
| Binary file detection + base64 | Streaming for very large files (>100 MB) |
| Atomic writes | File watching / change notifications |
| Glob pattern search | Full-text indexing |
| Regex find-and-replace | Multi-file transactional writes |
| Symlink resolution | Cross-platform symlink creation |
| MCP roots protocol | HTTP transport |
| Document text extraction / markdown conversion for formats supported by DocumentParsers | OCR for scanned documents |
See Also
Part of the AssistStudio ecosystem.
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.