Skip to main content
The Integrate.io MCP Server exposes a Model Context Protocol (MCP) endpoint that AI agents such as Claude, ChatGPT, Perplexity, and OpenAI Codex (or debugging tools like the MCP Inspector) can call to inspect and act on your account. Clients connect over JSON-RPC 2.0 and use a curated set of tools to list workspaces, list packages, inspect runs, search resources, validate pipelines, preview database data, and (with explicit mutating tools) clone packages, edit components, manage schedules, and run jobs. Credentials, TLS certificates, and SSL keys are never returned. For an overview of governance, version history and audit controls, see the Integrate.io MCP Server product page.

Server URL

Connect a client to the global URL: https://api.xplenty.com/api/mcp When you sign in, the consent screen asks which account to connect, and the connection is bound to that account. To connect several accounts at the same time, add one server per account with the per-account URL instead: https://api.xplenty.com/YOUR_ACCOUNT_ID/api/mcp Replace YOUR_ACCOUNT_ID with your account subdomain or numeric id (the value in your dashboard URL). See Connect a client for when to use each URL.
You don’t have to assemble these URLs by hand. Open Settings → Developer Settings → Connected Applications in your dashboard. The page shows the global URL first, labeled MCP server URL, and the per-account URL under Connect this account specifically (use this to connect several accounts at once).
Why api.xplenty.com instead of integrate.io? The Integrate.io API still runs at api.xplenty.com. This is expected and normal: Integrate.io was formerly called Xplenty, and the domain predates the rebrand. Your dashboard URL (dashboard.xplenty.com) and the API URL both work correctly.
The endpoint accepts a single JSON-RPC 2.0 message per request. Batch requests are not supported.

Authentication

Clients authenticate in one of two ways: an OAuth 2.1 browser sign-in (no token to copy or paste) or a static Personal Access Token sent as a bearer header. Which one you use depends on the client, and every detail (how to mint a token, how to revoke access, which clients are OAuth-first) lives on the Authentication page.

Prerequisites

  • Access: the Integrate.io MCP Server is available on every Integrate.io ETL & Reverse ETL account, with both read and write tools. There is nothing to enable.
    • Read tools inspect your account: inventory, search, schema discovery, previews, profiling, expression validation and docs Q&A.
    • Write tools build and edit packages, create and update schedules, and run, cancel or retry jobs.
  • A way to authenticate: either an OAuth-capable client or a Personal Access Token. See Authentication.
  • An MCP-capable client: see the support matrix below.

Get started

Quickstart

Connect Claude Code end to end in about five minutes.

Connect a client

Pick your client and follow its short setup page.

Client support matrix

Any MCP-capable client can connect. These are the clients we have verified, each with a short setup page.
MCP Inspector is not an AI agent. It is a debugging tool from the MCP project team (npx @modelcontextprotocol/inspector) that opens a local UI for listing and calling tools directly. Use it to confirm the endpoint works before wiring up a real agent.

What you can do with it

These are real prompts you can type into Claude (or any connected AI agent) once the MCP server is configured:

Tools

The server registers 49 tools across four areas: reading and inspecting, authoring and editing packages, jobs and schedules, and creating connections. Every tool is scoped to the authenticated account, obeys the signed-in user’s permissions, and never returns credentials, TLS certificates, or SSL keys. Read-only tools inspect account state; mutating tools create, modify, or run packages and jobs (or create connection links), and take effect immediately (there is no sandbox). Every tool result names the account it ran against, for example "account": {"name": "Acme", "subdomain": "acme"}, so you can tell connections to different accounts apart. Tool results also carry a call_id in their _meta; an agent that hits a problem can pass it to report_issue to tell the Integrate.io team. Run tools/list against the endpoint (for example from the MCP Inspector) to see the exact set your account exposes. Full parameter-level reference is split across four pages:

Reading & Inspecting

Read-only tools: inventory (connections, packages, runs, clusters), schema discovery, data preview and profiling, expression and package validation, documentation Q&A, and reporting issues back to Integrate.io.

Building & Editing Packages

Authoring tools: build a pipeline from intent, create or clone packages, add, edit, and remove components and edges, manage variables, save or discard a draft, and roll back versions.

Jobs & Schedules

Inspect and diagnose runs, count the records a run read and wrote, create and manage schedules, create a cluster, and run, cancel, or retry jobs.

Creating Connections

Create a database or file storage connection through a one-time dashboard link: you type the password or key on the page, and the assistant only learns the new connection’s ID.
Tools that hit external services (validate_package, validate_expression, discover_schema, discover_file_schema, preview_data, preview_transformation, profile_data, test_connection, run_package, ask_docs) make real network calls to JMS, the expression evaluator, the schema-importer, the previewer, the docs assistant, or your upstream systems. Use them sparingly.

Safety and audit

Mutating tools (build_pipeline, create_package, clone_package, update_package_components, update_package_edges, add_package_components, remove_package_components, rename_package, manage_package_variables, delete_package, revert_package, save_package_version, discard_draft, toggle_schedule, create_schedule, update_schedule, create_cluster, run_package, cancel_job, retry_job, create_connection_link) modify your live account, including production pipelines. There is no sandbox mode. Changes take effect immediately.
Audit trail: All writes made through the MCP server are captured in your account’s audit history via PaperTrail. Changes made by update_package_components are versioned and reversible through the existing package version history UI. Procurement checklist:
  • All MCP-driven changes appear in the PaperTrail audit log.
  • Tokens can be revoked instantly from Settings → Developer Settings → Connected Applications.
  • No credentials, TLS certificates, or SSL keys are ever returned by any tool.

Inline secret redaction in get_package

get_package(include_full_graph: true) returns a package’s full data_flow_json so an agent can clone or modify a pipeline by example. If component properties contain user-typed secrets, for example a REST source URL with ?api_key=..., an Authorization: Bearer ... header, a basic-auth password, or credentials embedded in a bash component body, those values are masked before the response leaves the server. The redaction is applied to the full-graph payload only. The summary returned by a default get_package call (counts and component-type tally) contains no secret-bearing fields and is unaffected. What gets masked
  • Values under keys whose name matches a secret pattern (password, token, api_key, client_secret, authorization, and similar).
  • The value side of { name, value } header pairs when the name side names a credential header (Authorization, X-API-Key, apikey, and similar). Non-credential headers like Content-Type are preserved.
  • URL userinfo credentials (https://user:pass@host).
  • URL query-string secret parameters (?api_key=..., &token=..., and similar).
  • Bearer and Basic tokens of 8 or more characters found anywhere in a string value.
Each masked value is replaced with the literal string [REDACTED]. Structure is preserved: component types, names, ids, positions, schema fields, edges, and non-secret properties (such as username) come through unchanged, so the canvas still renders and the agent can still reason about pipeline shape. Example A REST source with an api_key in the URL and an Authorization header comes back like this:
What this is and is not Redaction is a best-effort defense in depth for inline secrets that users sometimes paste directly into component properties. It is not a substitute for storing credentials in connections or in secret variables, which are encrypted at rest and never returned by any MCP tool. Use connections and secret variables for any real credential. Use these prompts to explore your account and test the integration. They use read-only tools only, so they are safe for production accounts.
list_workspaces is the best starting point for AI agents. It gives a snapshot of package counts and the latest run status of each workspace’s packages before diving into details.

Example: Calling a tool with curl

You can also call the endpoint directly. The body is a JSON-RPC 2.0 request.
To diagnose a failed run, call explain_job_failure with the run id:
To preview source data before designing a transformation:
To preview a single component’s output mid-build, before a destination is wired:

Authentication

Troubleshooting

Last modified on November 16, 2026