fd-mcp
Provides fast file search capabilities using fd (a modern find alternative), enabling AI assistants to efficiently navigate codebases, search file contents with ripgrep, and execute commands on matched files.
README
fd-mcp
A Model Context Protocol (MCP) server that provides fast file search capabilities using fd, a modern alternative to the traditional find command.
What is fd-mcp?
fd-mcp bridges fd with AI assistants like Claude Code through the Model Context Protocol. It exposes fd's powerful file search capabilities as MCP tools, enabling AI assistants to efficiently navigate and search codebases.
Why fd?
fd offers significant advantages over traditional file search tools:
- Blazing Fast: Written in Rust, fd is often 5-10x faster than
findfor typical searches - Recursive by Default: Searches directories recursively without special flags (use
max_depthto limit) - User-Friendly Syntax: Simple, intuitive patterns instead of cryptic flags (
fd patternvsfind -name "*pattern*") - Smart Defaults: Automatically respects
.gitignoreand skips hidden files/directories - Colorized Output: Enhanced readability with syntax highlighting
- Parallel Execution: Leverages multiple CPU cores for faster searches
- Regex Support: Built-in regex pattern matching without complex syntax
Use Cases
This MCP server is particularly useful for:
- Codebase Navigation: Quickly locate files by name, extension, or pattern across large projects
- Project Analysis: Find all files of a specific type (e.g., all Python test files, all configuration files)
- Code Exploration: Help AI assistants understand project structure by efficiently listing directories and files
- Pattern Matching: Search for files using regex patterns (e.g., find all migration files, test suites)
- Selective Searches: Filter by file type, depth, or exclude specific patterns
- Performance: Fast searches even in monorepos or large codebases with thousands of files
Prerequisites
Install fd:
# Ubuntu/Debian
sudo apt install fd-find
# macOS
brew install fd
# Arch
pacman -S fd
Install ripgrep (required for fd_search_content tool):
# Ubuntu/Debian
sudo apt install ripgrep
# macOS
brew install ripgrep
# Arch
pacman -S ripgrep
# Or download from: https://github.com/BurntSushi/ripgrep/releases
Installation
cd fd-mcp
pip install -e .
Usage with Claude Code
Add to your Claude Code MCP settings (~/.claude.json):
{
"mcpServers": {
"fd": {
"command": "fd-mcp"
}
}
}
Or run directly:
{
"mcpServers": {
"fd": {
"command": "python",
"args": ["-m", "fd_mcp.server"],
"cwd": "/home/th/dev/os/fd-mcp"
}
}
}
🚀 Claude Code Integration Best Practices
Once configured, Claude Code will have access to all fd-mcp tools. Here's how to get the most out of them:
Quick Start Commands
The project includes helpful slash commands to get started:
/find-py- Find all Python files (demonstratesfd_search)/search-code- Search code patterns (demonstratesfd_search_content)/recent- Show recently modified files (demonstratesfd_recent_files)/count-files- Count files by type (demonstratesfd_count)/demo-mcp- Run a full demonstration of all tools
Tool Selection Guide
When Claude Code needs to find files:
- ✅ Use:
mcp__fd__fd_search(pattern=".*", path=".", extension="py") - ❌ Instead of:
bash find . -name "*.py" - Why: 5-10x faster, respects .gitignore automatically
When Claude Code needs to search code:
- ✅ Use:
mcp__fd__fd_search_content(search_pattern="TODO", extension="py") - ❌ Instead of:
bash find . -exec grep "TODO" {} \; - Why: 10-100x faster, single operation, shows context
When Claude Code needs to find recent changes:
- ✅ Use:
mcp__fd__fd_recent_files(hours=24) - ❌ Instead of:
bash find . -mtime -1 - Why: Intuitive time params, faster execution
Performance Benefits
The tools are optimized for AI assistant workflows:
| Operation | Traditional | fd-mcp | Speedup |
|---|---|---|---|
| Find 1000 Python files | 2.5s | 0.3s | 8x faster |
| Search "TODO" in files | 15s | 0.5s | 30x faster |
| Find files changed today | 3s | 0.4s | 7x faster |
| Count all files | 2s | 0.25s | 8x faster |
Learning Resources
- See
CLAUDE.mdfor comprehensive usage patterns and examples - Check
.claude/commands/for ready-to-use slash commands - Review tool descriptions in Claude Code for quick reference
Tools
fd_search
Replaces: find, locate commands
Search for files and directories using fd (5-10x faster than find). Searches recursively by default through all subdirectories.
| Parameter | Type | Description |
|---|---|---|
| pattern | string | Regex pattern (required, use ".*" or "" for all) |
| path | string | Search directory (required, e.g., ".") |
| type | string | f=file, d=dir, l=symlink, x=exec, e=empty |
| extension | string | Filter by extension |
| hidden | bool | Include hidden files |
| no_ignore | bool | Don't respect .gitignore |
| max_depth | int | Max search depth |
| exclude | string | Glob pattern to exclude |
| case_sensitive | bool | Case-sensitive search |
| absolute_path | bool | Return absolute paths |
| max_results | int | Limit results (default: 100) |
fd_search_content ⭐
Replaces: find -exec grep, find | xargs grep commands
Search for content within files using fd+ripgrep. This is the key tool that replaces find . -exec grep pattern {} \; style commands.
| Parameter | Type | Description |
|---|---|---|
| search_pattern | string | Text/regex to search in files (required) |
| file_pattern | string | Filter files by name pattern |
| path | string | Search directory (default: ".") |
| extension | string | Filter by extension (e.g., "py", "js") |
| type | string | Filter by type (f=file, d=dir, etc.) |
| hidden | bool | Include hidden files |
| no_ignore | bool | Don't respect .gitignore |
| case_sensitive | bool | Case-sensitive search |
| context_lines | int | Lines of context around matches |
| max_results | int | Max files to search (default: 100) |
Note: Requires ripgrep (rg) to be installed.
fd_exec
Replaces: find -exec, find | xargs commands
Execute a command on files found by fd. Use {} as placeholder for filename.
| Parameter | Type | Description |
|---|---|---|
| command | string | Command to run (use {} for filename) |
| pattern | string | File name pattern |
| path | string | Search directory (default: ".") |
| type | string | Filter by type |
| extension | string | Filter by extension |
| hidden | bool | Include hidden files |
| no_ignore | bool | Don't respect .gitignore |
| max_files | int | Max files to process (default: 100) |
fd_recent_files
Replaces: find -mtime, find -newermt commands
Find recently modified files.
| Parameter | Type | Description |
|---|---|---|
| path | string | Search directory (default: ".") |
| hours | int | Modified within N hours (default: 24) |
| type | string | Filter by type |
| extension | string | Filter by extension |
| max_results | int | Limit results (default: 50) |
fd_count
Replaces: find | wc -l commands
Count files matching a pattern.
| Parameter | Type | Description |
|---|---|---|
| pattern | string | Regex pattern (required, use ".*" or "" for all) |
| path | string | Search directory (required, e.g., ".") |
| type | string | Filter by type |
| extension | string | Filter by extension |
| hidden | bool | Include hidden files |
Examples
Basic File Search
Find all Python files (recursively):
fd_search(pattern=".*", path=".", extension="py")
Find test files in entire project tree:
fd_search(pattern="test_.*", path=".", extension="py")
List directories only (limit to 2 levels deep):
fd_search(pattern=".*", path=".", type="d", max_depth=2)
Search only current directory (no recursion):
fd_search(pattern=".*", path=".", extension="py", max_depth=1)
Content Search (replaces find -exec grep)
Find "TODO" in all Python files:
fd_search_content(search_pattern="TODO", extension="py")
Find "import React" in JavaScript/TypeScript files with context:
fd_search_content(
search_pattern="import.*React",
extension="tsx",
context_lines=2
)
Find error handling in specific directory:
fd_search_content(
search_pattern="try.*except",
path="src/",
extension="py"
)
Execute Commands on Files
Count lines in all Python files:
fd_exec(command="wc -l {}", pattern=".*", path=".", extension="py")
Format all JavaScript files:
fd_exec(command="prettier --write {}", pattern=".*", path=".", extension="js")
Find Recent Changes
Files modified in last 2 hours:
fd_recent_files(hours=2)
Recent Python files modified in last day:
fd_recent_files(hours=24, extension="py")
Command Replacements
| Old Command | New MCP Tool |
|---|---|
find . -name "*.py" |
fd_search(pattern=".*", path=".", extension="py") |
find . -type f -exec grep "TODO" {} \; |
fd_search_content(search_pattern="TODO") |
find . -name "*.js" -exec prettier {} \; |
fd_exec(command="prettier {}", pattern=".*", path=".", extension="js") |
find . -mtime -1 |
fd_recent_files(hours=24) |
find . -type f | wc -l |
fd_count(pattern=".*", path=".", type="f") |
License
This project is licensed under the MIT License - see the LICENSE file for details.
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
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.
E2B
Using MCP to run code via e2b.