Dastur Docs
A powerful MCP server integration to bridge your AI agent with production-level context, environment & domain capabilities, low-level skills, and integrate seamlessly with your codebase.
Introduction
Dastur is an AI Skills Store — a hosted MCP server that supercharges your AI coding assistant with production-ready, domain-specific expertise. Instead of generic AI responses, get specialized skills built from real-world engineering experience.
Connect Dastur to Claude, Cursor, or any MCP-compatible AI tool and unlock capabilities like:
"Build this screen from my Figma design, integrate the user API, and make sure it follows our SwiftUI architecture patterns."
With Dastur skills enabled, your AI assistant understands platform conventions, generates production-quality code, and applies engineering best practices — turning hours of work into minutes.
Installation MCP
Dastur is a hosted MCP server that provides AI skills via the Model Context Protocol. Select your preferred AI tool below to get started.
Claude.ai Web Connector
EasiestThe quickest way to use Dastur. Connect directly in your browser with just a few clicks.
- Open Claude.ai in your browser
- Click the Settings icon (gear) in the bottom left
- Navigate to Connectors section
- Click Add Connector
- Enter the URL below and click Connect
https://api.growwithmugish.com/mcp
Claude Desktop App
RecommendedAdd Dastur to your Claude Desktop configuration for persistent access to all skills.
~/Library/Application Support/Claude/claude_desktop_config.json
(macOS)
%APPDATA%\Claude\claude_desktop_config.json
(Windows)
{
"mcpServers": {
"dastur": {
"url": "https://api.growwithmugish.com/mcp"
}
}
}
After saving, restart Claude Desktop:
# macOS
killall "Claude" && open -a "Claude"
# Windows
taskkill /IM "Claude.exe" /F && start "" "Claude"
Claude Code CLI
For DevelopersAdd Dastur directly from your terminal with a single command.
# Step 1: Navigate to your project directory
$ cd your-project
# Step 2: Start Claude Code
$ claude
# Step 3: Add Dastur MCP server
> claude mcp add -t http dastur https://api.growwithmugish.com/mcp
Added HTTP MCP server dastur with URL: https://api.growwithmugish.com/mcp to local config
Done. MCP server "dastur" added with URL https://api.growwithmugish.com/mcp.
# Step 4: Quit and restart Claude Code to apply changes
> /quit
$ claude
# Step 5: Verify Dastur is connected (type /mcp to check)
> /mcp
Manage MCP servers
1 server
> 1. dastur ✓ connected · Enter to view details
Shift+Tab on the first file prompt and select "Yes, allow all edits during this session" to auto-approve all file operations.
Cursor IDE
PopularIntegrate Dastur skills into Cursor's AI-powered coding environment.
~/.cursor/mcp.json
{
"mcpServers": {
"dastur": {
"url": "https://api.growwithmugish.com/mcp"
}
}
}
Or via Settings UI:
- Open Cursor Settings (
Cmd/Ctrl + ,) - Search for "MCP" or go to Features → MCP Servers
- Add new server with URL:
https://api.growwithmugish.com/mcp
Windsurf (Codeium)
NewAdd Dastur to Windsurf's MCP configuration for AI-enhanced development.
~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"dastur": {
"serverUrl": "https://api.growwithmugish.com/mcp"
}
}
}
ChatGPT (OpenAI)
Coming SoonChatGPT MCP support is currently in beta. Once available, you'll be able to connect Dastur skills directly.
MCP integration for ChatGPT is expected to be available soon. Stay tuned!
Notify me when availableGitHub Copilot
Coming SoonGitHub Copilot MCP support is on the roadmap. Integration details will be added once available.
We're working on GitHub Copilot integration. Check back soon!
Notify me when availableVerify Connection
Test that Dastur is connected by checking the health endpoint:
# Check server health
curl https://api.growwithmugish.com/health
# List available skills
curl https://api.growwithmugish.com/skills
Skills & Expertise
Skills define a comprehensive suite of specialized tools designed to accelerate the entire lifecycle of software engineering. From crafting pixel-perfect frontends, to architecting complex backend architectures, these skills provide the expertise your AI editor needs to become a full stack expert.
Frontend Engineering
iOS/SwiftUI Engineering
Enterprise iOS apps with clean MV/VM patterns and Core Data integration. Fully unit tested.
Android Native Engineering
Modern Android apps using Jetpack Compose, following MVVM clean arch with Hilt DI.
React Native Specialist
Cross-platform mobile with performance focus. Handles native modules and bridge optimization.
Modern Web Architecture
Server-side rendering, edge functions, and modern build pipelines for blazing fast web apps.
Backend Systems
High-Performance API Systems
Build high-throughput REST & GraphQL APIs. Expert in rate limiting, caching, and async patterns.
Database Engineering
Schema design, query optimization, replication setup, and data migration strategies.
Microservices Architecture
Design distributed systems with service mesh, message queues, and observability built-in.
Serverless Computing
Event-driven architectures with AWS Lambda, Cloudflare Workers, and edge computing patterns.
DevOps & Infrastructure
Container Orchestration
Manage large-scale container deployment to get any container up and running on your infrastructure.
Full-Stack Observability
Comprehensive monitoring, logging, and tracing for production systems with alerting.
Automated Pipelines
Build, test, deploy pipelines with GitHub Actions, GitLab CI, and Jenkins. Full automation.
Infrastructure as Code
Provision and manage cloud resources with Terraform, AWS CDK, and infrastructure automation.
Adding Your Own Skills Coming Soon
Extend Dastur's capabilities by adding custom skills to make resource discovery with nothing for now. Below is a simple example:
# skills/my_custom_skill.py
from dastur.skill import Skill, tool
class MyCustomSkill(Skill):
name = "my-custom-skill"
description = "A custom skill for specialized tasks"
@tool
def analyze_code(self, file_path: str) -> dict:
"""Analyze a code file and return insights."""
# Your implementation here
return {"status": "analyzed"}
Ready to superpower your workflow?
Join the Dastur community to share skills, request features, and get support directly from the dev.