Connect Instagram DM Automation to Claude & ChatGPT via MCP (2026)

What Inrō's MCP server actually does

Inrō's MCP server is a live connection between your Instagram DM automation account and an AI agent like Claude or ChatGPT. Once connected, your AI agent can read everything in your Inrō account and take actions on your behalf, using plain language, with no need to navigate the platform manually.

This matters for two reasons. First, most Instagram automation users have more data in their account than they can easily read. Contacts accumulating across folders, campaigns sent and forgotten, scenarios running in the background with no clear picture of what is working. The MCP connection makes that data queryable. Second, a lot of account management tasks are repetitive and manual: moving contacts, creating folders, setting up campaigns, triggering scenarios. The MCP connection makes those tasks delegatable.

Inrō's MCP server gives your AI agent access to 25 tools across six areas of your account.

Use it to understand what's happening in your account

The first thing most automation users want when they connect their account to an AI agent is clarity. Not strategy advice. Not generic tips. Just a clear picture of what is actually happening.

Inrō's MCP server makes your interaction data queryable in plain language. Here is what that looks like in practice:

Understanding your contacts:

• "Show me all contacts who entered my lead magnet scenario this week."
• "Where is [contact name] in my funnel? Which folder are they in and what did they receive?"
• "Find contacts who replied to my last campaign but haven't been added to a follow-up folder."
• "Search for contacts who came in through story replies in the last 30 days."

Understanding your campaigns:

• "What were the open rates on my last five campaigns? Which performed best?"
• "Which campaign had the most link clicks? What was the message?"
• "Show me everyone who received my March campaign and didn't click."

Understanding your scenarios:

• "List all my active scenarios. Which ones haven't triggered in the last 30 days?"
• "Which scenario has triggered the most times this month?"

These answers exist in your Inrō account right now. The MCP connection means you can get them in seconds, in the same conversation where you decide what to do next.

Use it to delegate account tasks to your AI agent

The second use case is less about understanding and more about doing. Inrō's MCP server gives your AI agent write access to your account, which means it can execute tasks on your behalf once you approve them.

Managing contacts and folders:

• "Create a folder called 'Hot Leads April' and move these 15 contacts into it."
• "Find all contacts tagged as 'interested' who aren't in any campaign yet and add them to my nurture folder."
• "Update the contact record for [name] to add the property 'VIP'."

Setting up and sending campaigns:

• "Create a new DM campaign for my April launch targeting contacts in my 'warm leads' folder."
• "Send my re-engagement campaign to everyone who opened but didn't click my last message."
• "Update the copy on my last campaign draft before I send it."

Working with scenarios:

• "I want to send a freebie when someone comments 'yes' on my next Reel. Which template in Inrō fits this?"
• "Trigger my welcome scenario for this new contact."
• "Create a new scenario based on [template name]."

For scenario creation, the MCP server works with Inrō's existing template library: your AI agent can recommend the right template and create a scenario from it. If your use case goes beyond what the library covers, Inrō's built-in scenario builder accepts a plain-language description and generates a fully custom automation directly inside the platform.

Sending individual messages:

• "Send a DM to [contact name] with this message."
• "Message everyone in my 'no reply' folder with this follow-up."

Write actions default to "Needs approval" in Claude's permission settings. Your AI agent will always show you what it is about to do before executing.

How agencies use the MCP connection at scale

For agencies managing Instagram automation across multiple client accounts, the combination of data querying and task delegation compresses hours of account management into a single conversation.

A typical audit that previously required opening each client account manually now looks like this in Claude:

• "Find all contacts across my folders who haven't been messaged in 14 days."
• "Which of my active scenarios have triggered fewer than 10 times? List them."
• "Create a re-engagement campaign for the 'cold leads' folder and draft the message."

Inrō executes each action. The account manager reviews and approves. Each client organisation requires a separate MCP connection: switch accounts in Inrō, re-authorise the connector, and the agent has full access to that client's data.

For a broader view of how Inrō connects to your existing tools and workflows, see Inrō's integrations overview.

How to connect Inrō to Claude or ChatGPT

Setup for Claude takes under 5 minutes and requires no technical knowledge.

1. Copy your MCP Server URL. In Inrō, go to your account menu (bottom-left), open Integrations & API, find the MCP Server card, and click Copy. The URL is https://api.inro.social/mcp.
2. Open Claude's connector settings. Go to claude.ai, click your profile icon, then go to Settings > Connectors. Click Add custom connector.
3. Paste the URL and name the connector. Enter the URL, name it "Inrō", and click Add.
4. Authorize access. Click Connect, review the Inrō authorization screen, and click Authorize.
5. Set tool permissions. Read tools like list_contacts, get_campaign, and list_scenarios can be set to "Always allow." Keep write tools like send_message, send_campaign, and trigger_scenario on "Needs approval."
6. Open a new chat and ask your first question.

For ChatGPT, setup uses Developer Mode (available on Plus, Business, Enterprise, and Edu plans). OAuth endpoints are auto-discovered from Inrō's server, so no manual configuration is needed. ChatGPT's MCP support is currently in beta.

Set up Inrō's MCP connection in under 5 minutes.

Connect your automations to an AI agent that can actually read it, query it, and manage it for you.

Start for free

FAQs

What is Inrō's MCP server for Instagram DM automation?

‍Inrō's MCP server is a connection that gives AI agents like Claude and ChatGPT live access to your Instagram DM automation account. Through 25+ tools, your AI agent can query contacts, campaigns, scenarios, folders, and conversation history in plain language. It runs through Meta's official API and is the first MCP server built specifically for Instagram automation.

What kind of interaction data can I query through the MCP connection?

‍You can ask about any interaction stored in Inrō: who commented and triggered a scenario, who DMed you or received a DM from you, which contacts opened or clicked a campaign, what folder a contact is in, and the full automation history for any individual contact. The MCP server pulls live data, not estimates.

Can my AI agent create or modify scenarios?

‍The MCP server gives your AI agent access to Inrō's template library and can recommend the right scenario for your use case. It cannot generate custom scenarios from a free-text prompt. For fully custom automations, Inrō's built-in scenario builder accepts a plain-language description and generates the scenario directly inside the platform.

Is it safe to give Claude or ChatGPT access to my automation account?

‍Yes. Inrō's MCP server runs through Meta's official API. Write actions require your approval by default through Claude's per-tool permission controls. There is no sandbox mode, so reviewing write actions before confirming is recommended.

What is the difference between Inrō's MCP server and a generic Instagram MCP tool?

‍Generic Instagram MCP tools connect to Instagram's native API and surface post, comment, and basic profile data. Inrō's MCP server exposes your full automation stack: contact history, campaign performance, scenario library, CRM folders, and DM conversation records. This is the data layer that matters for anyone running Instagram automation at scale.

Do I need coding skills to set up the connection?

‍No. Setup for Claude requires pasting a URL into the custom connector settings and clicking Authorize. ChatGPT's beta setup takes a few more steps but is also entirely no-code.

Can agencies use this across multiple client accounts?

‍Yes. Each Inrō organisation requires a separate MCP connection. Switch accounts in Inrō, re-authorise the connector in Claude or ChatGPT, and the agent has access to that client's data. There is no limit on the number of organisations you can connect across separate sessions.