Intercom MCP Integration

Weldable's Intercom MCP integration connects your AI agents to Intercom's customer messaging platform for managing conversations, contacts, and support workflows. Intercom has gone all-in on AI-first support in 2026, with its Fin AI Agent participating in 99% of conversations at companies like Lightspeed Commerce and autonomously resolving up to 65% of them. Weldable gives your agent direct access to Intercom's API so you can build custom automations that work alongside Fin and your human support team.

Use cases

Escalation routing with context

When a customer conversation in Intercom reaches a complexity threshold, your agent extracts the full conversation history, pulls related account data from your CRM or database, and creates a structured escalation summary. This summary can be posted to a Slack channel for the engineering team or assigned to a specific support agent with all the context they need to resolve the issue quickly.

Customer health monitoring

Your agent periodically queries Intercom for conversation trends by customer account, tracking metrics like response times, conversation frequency, and unresolved issues. It compiles a health report and posts it to Slack or Google Sheets, flagging accounts that show signs of frustration or churn risk before they escalate.

Proactive outreach based on product events

When your product logs a significant event, such as a failed payment, a feature adoption milestone, or an approaching subscription renewal, your agent creates a targeted message in Intercom for the affected user. The message is personalized with account-specific details rather than a generic template.

Support ticket analysis

Your agent pulls closed conversations from Intercom over a given time period, categorizes them by topic and resolution type, and generates a report highlighting the most common issues. This data feeds into product decisions by showing where customers struggle most, without requiring manual tagging of every conversation.

Knowledge base gap detection

Your agent reviews recent Intercom conversations where the AI could not resolve the issue autonomously, identifies the questions that lacked help center coverage, and creates a prioritized list of articles to write. This closes the loop between support interactions and documentation.

How it works

Connect your Intercom workspace through OAuth. Weldable requests scopes for conversation management, contact access, and messaging. Tokens stay current through automatic refresh, so your agent maintains access without manual re-authentication.

Describe what you need and Weldable routes your intent to the appropriate Intercom API endpoint. Your agent can search conversations, create messages, update contact attributes, and manage tags. These actions chain with other integrations, so a single workflow can read from Intercom and act on the data in Slack, Google Sheets, or any other connected service.

Tips

Use tags for automated categorization. When your agent processes conversations, applying consistent tags makes it possible to filter and report on conversation types later. Establish a tagging taxonomy and let the agent apply it as part of every workflow.

Respect Intercom's rate limits on conversation queries. Intercom throttles API requests, particularly on endpoints that return large conversation lists. Have your agent paginate through results and space out bulk queries rather than requesting everything at once.

Combine with Slack for internal visibility. Your agent can post conversation summaries or escalation alerts to a dedicated Slack channel, giving the support team a real-time view of what is happening in Intercom without everyone needing to monitor the inbox.

Pull conversation data for product feedback loops. The richest source of product feedback often lives in support conversations. Your agent can extract themes from resolved tickets and present them in a format that product teams can act on.

Keep outreach messages concise and specific. When your agent sends proactive messages through Intercom, shorter messages with a clear purpose get better response rates than long explanations. State the reason for reaching out and the next step.