Connect Your AI Assistant to Oshyn's Free DXP Tools via MCP A step-by-step guide to adding Oshyn's MCP server to Claude, ChatGPT, Cursor, and more.

Oshyn's suite of free DXP tools is now available as a single Model Context Protocol (MCP) server, making it easier to bring Oshyn's digital experience strategy tools directly into the AI assistants and development environments you already use.

Connecting your AI assistant to the MCP server lets you ask questions in plain language and have your assistant call the right Oshyn tool in the background. It can help identify gaps in how AI sees your brand, evaluate site reliability, explore DXP options, estimate project budgets, get Oshyn insights, and guide planning conversations without leaving your chat.

This guide is for marketers, digital leaders, developers, and AI power users who want faster access to practical DXP guidance from inside Claude, ChatGPT, Cursor, and other MCP-compatible tools.

With one connection, you can:

• Use Oshyn's tools conversationally, with no forms to fill out and no tabs to switch between.
• Access a single server URL that exposes 31 operations across Oshyn's free DXP tools.
• Connect once, then query whenever you like with your preferred AI assistant.

Start by adding the MCP server URL below, then use the platform-specific instructions in this guide to get connected.

MCP Server URL: https://mcp.oshyn.com

What You Can Do Once Connected

After connecting, you can ask your AI assistant things like:

• "Run a site reliability report on acme.com — my email is jane@acme.com"
• "Help me figure out which DXP platform is right for our organization"
• "Estimate the budget for a Sitecore-to-headless migration with 300 pages"
• "I want advice on whether we need a full redesign or a migration"
• "Search the Oshyn blog for posts about AEM Edge Delivery Services"
• "Show me Digital Trust Index scores for the financial services sector since January 2026"

The assistant handles the tool selection and parameter passing automatically — you just have a conversation.

Compatibility Overview

We've tested the MCP server across the major AI assistants and IDEs. Here's what works natively:

Platform	Works?
Claude
Claude Code / Claude CLI
ChatGPT
OpenAI Codex IDE / CLI
Cursor
Devin Desktop
xAI Grok
Perplexity
Mistral (Le Chat)
Amazon Q Developer
GitHub Copilot	Via mcp.json
Microsoft Copilot	Via mcp.json
Gemini IDE extension	Via mcp.json
Gemini Chat
Ollama (local)	Via bridge client
DeepSeek	Via API + client
LLaMA (Meta AI Chat)	Via local MCP client

Installation Instructions by Platform

Claude (claude.ai, Web, Mobile, Desktop)

Available on Pro, Max, Team, and Enterprise plans.

1. Go to claude.ai/customize/connectors
2. Click the + icon, then Add custom connector
3. Enter the MCP URL: https://mcp.oshyn.com
4. Give it a name (e.g., Oshyn Tools) and save

That's it. Oshyn's tools will now be available in every new Claude conversation.

Claude Code / Claude CLI / Claude Cowork

You have three scope options depending on how broadly you want the server available:

Globally (available in all projects):

claude mcp add oshyn -s user --transport http --url https://mcp.oshyn.com

Saved to ~/.claude.json

Per project (committed to source control):

claude mcp add oshyn -s project --transport http --url https://mcp.oshyn.com

Saved to mcp.json at the root of your project

Local only (not committed to source control):

claude mcp add oshyn -s local --transport http --url https://mcp.oshyn.com

Saved to .claude/settings.local.json

ChatGPT

Available on Plus, Pro, and Business plans. 1 Go to chatgpt.com/\#settings/Connectors 2 Click Advanced Settings and enable Developer Settings 3 Click Create App 4 Enter https://mcp.oshyn.com as the MCP URL and give it a name 5 Save

OpenAI Codex (IDE)

Via the Settings UI: 1 Open Settings → MCP Servers 2 Click + Add Server 3 Enter https://mcp.oshyn.com and a name, then save

Via the command line:

| # Direct HTTP (preferred)codex mcp add oshyn --url https://mcp.oshyn.com# Via mcp-remote if direct HTTP isn't availablecodex mcp add oshyn -- npx -y mcp-remote `https://mcp.oshyn.com`# Verify it was addedcodex mcp list |

Cursor

1 Open the MCPs menu in Cursor 2 Click the + symbol → Add custom MCP 3 Enter https://mcp.oshyn.com and a name 4 Save

Devin Desktop

1 Click File → Preferences → Devin Settings 2 Under Devin CLI, locate MCP Servers 3 Enter https://mcp.oshyn.com 4 Click Add

xAI Grok

Available on paid plans. The server must be reachable from the public internet. 1 Open Settings → Connectors 2 Click New Connector → Custom 3 Paste https://mcp.oshyn.com, give it a name 4 Save — Grok will automatically pull the available tools

Perplexity

Available on Pro, Max, and Enterprise plans. 1 Go to Settings → Connectors 2 Click Add custom connector 3 Enter https://mcp.oshyn.com and a name 4 Choose your authentication method (none required for the Oshyn server) and save

Mistral (Le Chat)

Available on paid plans. 1 In Le Chat, open Connectors → Custom MCP Connector 2 Enter a name (no spaces) and paste https://mcp.oshyn.com 3 Set authentication to none (no auth required) 4 Click Connect

Amazon Q Developer

Via mcp.json (global):

Add the following to ~/.aws/amazonq/mcp.json:

{
  "mcpServers": {
    "oshyn": {
      "type": "http",
      "url": "https://mcp.oshyn.com/"
    }
  }
}

For a single workspace, add the same block to .amazonq/mcp.json at the project root. Via CLI:

q mcp add oshyn --type http --url https://mcp.oshyn.com/

The server will be picked up the next time you start a Q Developer chat session.

GitHub Copilot, Microsoft Copilot, and Gemini IDE

These environments don't support custom remote MCP connectors through a UI, but you can use the mcp.json file approach in your workspace:

{
  "mcpServers": {
    "oshyn": {
      "type": "http",
      "url": "https://mcp.oshyn.com/"
    }
  }
}

Place this file at the root of your project or in the appropriate configuration directory for your IDE.

Local / Self-Hosted Models (Ollama, LLaMA, DeepSeek)

These models don't ship with a built-in MCP client, but you can connect them via a bridge or MCP-capable client wrapper:

• Ollama: Use a bridge like ollmcp, MCPHost, or ollama-mcp-bridge. Point the bridge at https://mcp.oshyn.com, and it will surface the tools as standard Ollama tool calls.
• LLaMA (Meta AI): Run through a Mac client that supports MCP, such as LM Studio, and add the server to that client's mcp.json.
• DeepSeek: The DeepSeek app doesn't have a connector screen. Run the model through an agent studio or any MCP-capable API client and point it at https://mcp.oshyn.com.

MCP Capabilities Overview

The server provides access to the tools below. You interact with these tools using natural language; the AI assistant automatically selects the appropriate tool and manages the required parameters.

Why MCP?

The Model Context Protocol is an open standard for connecting AI assistants to external tools and data sources. Rather than copying and pasting between tools, or asking your AI assistant to guess at information it doesn't have, MCP lets the assistant call live tools directly — in the middle of a conversation, with real data.

For Oshyn's tools, that means your AI assistant can run an actual site reliability scan, not just describe what one looks like. It can walk you through the DXP MatchMaker questionnaire and return a real platform recommendation, not a generic answer about CMS selection.

Oshyn's 25 years of DXP implementation expertise are now available to query from wherever you already work. Questions about the MCP server or Oshyn's tools? Get in touch with us — we're happy to help you get connected.