Back to Tools
Customer Support API Key
customerio logo

Customer.io MCP Tools for AI Agents

Let an agent find people, inspect profiles and message history, record events, trigger broadcasts, and send transactional email in Customer.io.

Available Tools

18 tools available in this integration

Customerio.DeletePerson

Delete a person record to honor a data-removal request.

Customerio.GetCampaignMetrics

Read a campaign's aggregate delivery metrics over a period.

Customerio.GetNewsletterMetrics

Read a newsletter's aggregate delivery metrics over a period.

Customerio.GetPerson

Look up a person's profile by email or id.

Customerio.IdentifyPerson

Create or update a person and their attributes.

Customerio.ListBroadcasts

List the workspace's broadcasts to discover ids that can be triggered.

Customerio.ListCampaigns

List the campaigns in the workspace.

Customerio.ListNewsletters

List the newsletters in the workspace.

Customerio.ListPersonActivities

List a person's recent activities, most recent first, resolved by email

Customerio.ListPersonMessages

List a person's message and delivery history, most recent first,

Customerio.ListPersonSegments

List the segments a person belongs to, resolved by email or id.

Customerio.ListSegmentMembers

List the ids of the people in a segment and the segment's total size.

Customerio.ListSegments

List the segments defined in the workspace.

Customerio.ListTransactionalMessages

List the transactional message templates in the workspace.

Customerio.RecordEvent

Record a named custom event on a person's timeline to drive lifecycle

Customerio.SearchPeople

Find people by segment membership and/or an attribute, to build an

Customerio.SendTransactionalEmail

Send a transactional email to one recipient by its template id.

Customerio.TriggerBroadcast

Trigger an API broadcast to a segment or an explicit recipient list.

Working directly against the raw API? 139 more tools ship across 3 raw-API servers.

Why agent-optimized tools matter

Built for intent, not API schemas

Ask the agent to "find a customer by email, check their recent messages, and trigger the right broadcast." Arcade helps it use the right Customer.io tools without guessing person, segment, campaign, or broadcast IDs.

Up to 100x fewer tokens

Customer.io list and search operations return focused people, segment, campaign, and message records instead of entire workspace exports. The agent can inspect activity and delivery history only for the people it needs.

Complete the workflow end to end

Find a person, inspect their segments and message history, record an event, request approval, and trigger a broadcast through one toolkit designed around the complete customer messaging task.

Make actions predictable and controlled

Arcade tools return structured results and useful errors across users and runs. Teams can apply scope restrictions, redaction, and approval policies so agents follow the same controls every time.

Securing Customer.io

The more capabilities agents get, the more useful they are. However, more capabilities bring more risks of undesired effects such as data loss and exfiltration.

This is why Arcade provides you with contextual rules that dictate how and when a tool is allowed to run. Arcade’s contextual access lets you govern tool visibility and behavior at every stage of execution and control who can see a tool, what inputs are allowed, and what comes back.

Learn about Contextual Access Policies

For Customer.io, you can implement some of these policies:

Access when tools are listed for a user
  • Hide Customerio.DeletePerson from agents unless the roles your IDP marks as privacy administrators are active, so an agent cannot erase a customer record without the right authority.
  • Restrict Customerio.SendTransactionalEmail to agents operating under your support or communications role, so unapproved workflows cannot send messages to individual customers.
  • Deny Customerio.TriggerBroadcast to agents outside the roles your IDP marks as lifecycle owners, so an agent cannot send a broadcast to a segment or recipient list without campaign authority.
Pre-execution before each tool call
  • Require human approval before Customerio.IdentifyPerson runs when it changes a person's email, consent status, or other contact attribute, so an agent cannot overwrite customer identity or preferences without review.
  • Block Customerio.RecordEvent when the event name or attributes contain credentials, payment details, or message content, so an agent cannot place sensitive data on a customer's activity timeline.
  • Require human approval before Customerio.DeletePerson runs, so a mistaken match or automated data-removal request cannot permanently erase the wrong customer record.
Post-execution after the tool returns
  • Redact Customerio.GetPerson results before they reach the model, stripping phone numbers, postal addresses, and consent metadata unless the agent is handling an approved customer-data request.
  • Strip message bodies, unsubscribe links, and recipient addresses from Customerio.ListPersonMessages results, so the model sees delivery history without exposing private customer communications.
  • Redact email addresses, phone numbers, and postal addresses from Customerio.SearchPeople results, so a segment lookup returns matching records without revealing direct contact details.

How to use Customer.io with your agent

Add your Customer.io API key once, then connect an MCP client to let an agent work with your workspace. Pick your client below and follow the setup from start to finish.

Your questions, answered

A Customer.io MCP server exposes Customer.io capabilities as structured tools that an AI agent can call through the Model Context Protocol. Agents can look up people, inspect message history and segments, update profiles, record events, and send or trigger messages within the configured credential's access.

Arcade provides an agent-optimized Customer.io integration and adds the runtime around it: shared-key authentication, approval policies, response redaction, tool access controls, and one MCP connection for supported clients.

Arcade agents can look up people, search by attributes or segment membership, inspect activities, messages and segments, list campaigns and newsletters, read delivery metrics, update profiles, record events, delete person records, send transactional emails, and trigger broadcasts.

Configure one Customer.io API key or secret with access to the Customer.io data and actions your workflow needs. That includes person records, events, broadcasts, transactional messages, and deletion when those operations are enabled; use a credential with the narrowest access that supports the job.

No. Customer.io uses a shared API key or secret configured once for the integration, rather than having every user authorize a separate account. Arcade brokers that credential, and the model never receives it.

Yes. Contextual Access policies can require human approval before updating or deleting a person, recording an event, sending a transactional email, or triggering a broadcast. The call proceeds only after the policy's conditions have been satisfied.

Arcade proxies Customer.io API calls and does not persist your content as part of normal tool execution. Credentials are brokered and stored encrypted at rest, and the model never receives them.

Any framework or client that can connect to Arcade's MCP gateway or tool APIs works, including LangChain, OpenAI Agents, CrewAI, Claude Desktop, Claude Code, Cursor, Visual Studio Code, Microsoft Copilot Studio, and GitHub Copilot.

Arcade refreshes access while the Customer.io credential remains valid. If the key is revoked, rotated, expires, or can no longer be used, calls stop until the credential is reconnected or replaced.

Customer.io's API rate limits and account or workspace quotas still apply to calls made through Arcade. The applicable limits depend on your Customer.io plan and API access, so high-volume searches, event recording, and message sends should follow Customer.io's current limits.

Ready to actually ship your agent?

Connect your agent to Customer.io with a shared API key, agent-optimized tools and approval controls for consequential actions.