OODA Computer Control
Provides comprehensive computer control capabilities including shell commands, file operations, screen capture, keyboard/mouse input, window management, clipboard access, and system monitoring following the OODA loop pattern.
README
mnehmos.ooda.mcp (v2.0.0)
A comprehensive MCP (Model Context Protocol) server that provides full computer control capabilities to Claude. Implements the OODA loop (Observe-Orient-Decide-Act) pattern for autonomous computer interaction.
Features
| Category | Tools | Description |
|---|---|---|
| CLI & Files | 17 | Shell commands, file read/write/copy/move/delete, search |
| CRUD Database | 9 | Persistent SQLite key-value store with collections |
| Screen (Observe) | 4 | Screenshot, display info, screen change detection |
| Input (Act) | 10 | Keyboard typing/shortcuts, mouse move/click/drag/scroll |
| Window Management | 10 | List/focus/minimize/maximize/close windows, launch apps |
| Clipboard | 4 | Read/write text, HTML, images |
| System | 8 | System info, processes, network, notifications |
Total: 62 tools with batch/parallel execution support for most operations.
Security Warning
This server provides unrestricted system access. Claude will be able to:
- Execute arbitrary shell commands
- Control keyboard and mouse
- Take screenshots
- Read/write files anywhere
- Manage windows and processes
Use at your own risk in trusted environments only.
Installation
git clone https://github.com/Mnehmos/mnehmos.ooda.mcp.git
cd mnehmos.ooda.mcp
npm install
npm run build
Claude Desktop Configuration
Add to your Claude Desktop config file:
Windows: %APPDATA%\Claude\claude_desktop_config.json
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"ooda-computer": {
"command": "node",
"args": ["/absolute/path/to/mcp-ooda-computer/dist/index.js"]
}
}
}
OODA Loop Pattern
The tools are designed around the OODA (Observe-Orient-Decide-Act) loop:
Observe
screenshot- Capture screen or region (returns base64 or saves to file)get_screen_info- Get display/monitor informationwait_for_screen_change- Detect when screen content changeslist_windows- See all open applicationsget_active_window- Know current focusclipboard_read- Check clipboard contents
Orient
- Use CRUD tools to store/retrieve context and state
file_info,search_files- Understand filesystem statelist_processes,get_system_info- Understand system state
Decide
- Built into Claude's reasoning between observations
- CRUD database for persistent decision context
Act
keyboard_type,keyboard_shortcut- Type and use hotkeysmouse_click,mouse_move,mouse_drag- Mouse controlfocus_window,launch_application- Window controlclipboard_write- Set clipboard contentnotify- System notifications
API Reference
CLI & File Operations
| Tool | Description |
|---|---|
exec_cli |
Execute shell command |
read_file |
Read file contents |
write_file |
Write content to file |
list_directory |
List directory contents |
copy_file |
Copy file/directory |
move_file |
Move/rename file |
delete_file |
Delete file/directory |
file_info |
Get file metadata |
search_files |
Search files by pattern |
batch_* |
Parallel versions of above |
CRUD Operations
| Tool | Description |
|---|---|
crud_create |
Create record in collection |
crud_read |
Read record by ID |
crud_update |
Update existing record |
crud_delete |
Delete record |
crud_query |
Query with filters |
crud_batch_* |
Parallel versions |
Screen Operations
| Tool | Description |
|---|---|
screenshot |
Capture screen (region optional) |
get_screen_info |
Display/monitor info |
wait_for_screen_change |
Wait for UI changes |
find_on_screen |
Find text/image (requires OCR) |
Input Operations
| Tool | Description |
|---|---|
keyboard_type |
Type text |
keyboard_press |
Press key with modifiers |
keyboard_shortcut |
Execute shortcut (e.g., "ctrl+c") |
mouse_move |
Move cursor |
mouse_click |
Click at position |
mouse_drag |
Drag between points |
mouse_scroll |
Scroll wheel |
get_mouse_position |
Current cursor position |
batch_keyboard_actions |
Sequence of keyboard actions |
batch_mouse_actions |
Sequence of mouse actions |
Window Operations
| Tool | Description |
|---|---|
list_windows |
All open windows |
get_active_window |
Currently focused window |
focus_window |
Bring window to front |
minimize_window |
Minimize window(s) |
maximize_window |
Maximize window |
restore_window |
Restore from min/max |
close_window |
Close window |
resize_window |
Resize window |
move_window |
Move window position |
launch_application |
Start application |
Clipboard Operations
| Tool | Description |
|---|---|
clipboard_read |
Read text/HTML/image |
clipboard_write |
Write text/HTML |
clipboard_clear |
Clear clipboard |
clipboard_has_format |
Check format availability |
System Operations
| Tool | Description |
|---|---|
get_system_info |
OS, CPU, memory, uptime |
list_processes |
Running processes |
kill_process |
Kill by PID or name |
get_environment |
Environment variables |
set_environment |
Set environment variable |
get_network_info |
Network interfaces |
wait |
Sleep for milliseconds |
notify |
System notification |
Batch Operations
Most tools have batch versions for parallel execution:
batch_exec_cli - Multiple commands in parallel
batch_read_files - Multiple files in parallel
batch_write_files - Multiple files in parallel
batch_copy_files - Multiple copies in parallel
crud_batch_create - Multiple records in parallel
...
Batch operations return structured results:
{
"summary": { "total": 5, "successful": 4, "failed": 1, "elapsed_ms": 23 },
"results": [
{ "index": 0, "success": true, "result": {...} },
{ "index": 1, "success": false, "error": "..." }
]
}
Platform Support
| Feature | Windows | macOS | Linux |
|---|---|---|---|
| CLI/Files | ✅ | ✅ | ✅ |
| CRUD | ✅ | ✅ | ✅ |
| Screenshot | ✅ PowerShell | ✅ screencapture | ✅ scrot |
| Keyboard | ✅ SendKeys | ✅ osascript | ✅ xdotool |
| Mouse | ✅ user32.dll | ⚠️ Limited | ✅ xdotool |
| Windows | ✅ user32.dll | ✅ osascript | ✅ wmctrl |
| Clipboard | ✅ PowerShell | ✅ pbcopy/paste | ✅ xclip |
Configuration
Optional config file at ~/.mcp/config.json:
{
"storage": {
"type": "sqlite",
"path": "~/.mcp/workspace.db"
},
"cliPolicy": {
"mode": "allow-all",
"extraBlockedPatterns": [],
"timeoutMs": 30000
},
"crud": {
"defaultLimit": 1000
}
}
Development
npm install # Install dependencies
npm run build # Build TypeScript
npm run dev # Watch mode
npm start # Run server
Known Limitations
find_on_screenrequires additional OCR dependencies (not included)- macOS mouse position reading returns 0,0 (system limitation)
- Some mouse drag operations may require platform-specific tuning
License
MIT License - see LICENSE file.
Changelog
See CHANGELOG.md for version history.
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.
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.