Developers Hub

Build on Next Gate Tech

Use APIs, webhooks, MCP, and in-platform code to integrate and extend the platform.

Request Access

API & MCP

Powerful API & MCP for Developers

Build custom integrations and workflows with our comprehensive REST API and Model Context Protocol (MCP) server.

Connect AI agents to your data

Configure your MCP-compatible AI tools to connect directly to the platform. Query data, manage workflows, and automate operations through Model Context Protocol.

View Docs
jsonc
// Add to your MCP client config
{
  "mcpServers": {
    "nextgatetech": {
      "command": "npx",
      "args": ["-y", "@nextgatetech/mcp-server"],
      "env": {
        "NEXTGATETECH_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Access

Two Doors In, One Permission Model

The dashboard and your integrations call the REST API. Agents call the MCP server. Both endpoints sit on the same host and pass through the same middleware, so a call gets exactly the same permissions whichever door it came through.

REST API

apis.nextgatetech.com. Bearer tokens. What the dashboard itself uses, so anything you see there you can automate.

MCP server

apis.nextgatetech.com/mcp. OAuth or bearer tokens. Claude Desktop, ChatGPT or any MCP-capable client sees SPARK as a set of tools.

Shared auth middleware

Read the token, validate it with your identity provider, identify the caller, check the caller's permissions. Four steps, one implementation, both doors.

One permission model

A person in the dashboard and an agent over MCP acting for the same role get the same answer. Permissions are set once, in Governance.

Experimentation

Playground

Small experiments and fun builds from the engineering team, exploring ideas beyond the day-to-day.

REST API

Build with our REST API

Authenticate, read, write, and control workflows programmatically.

OAuth 2.0 authentication

Authenticate requests and scope what each client can do.

Data read operations

Retrieve Next Gate Tech data for your own reporting, checks, and downstream processing.

Data write operations

Create or update Next Gate Tech data from your upstream sources.

Workflow control

Start, stop, and manage workflow runs from external systems.

Workflow state queries

Read current workflow status to coordinate handoffs and monitoring.

Custom connectors

Build custom connectors that use the API to extend your integration coverage.

MCP Server

Connect tools and agents through the MCP server

Connect your MCP client to the Next Gate Tech MCP server to work with data and workflows.

MCP client connection

Point your MCP client at the Next Gate Tech MCP server.

Webhooks

Receive events with webhooks

Let Next Gate Tech call your endpoints when data or workflow state changes.
Events

Data change notifications

Receive event payloads when Next Gate Tech data is updated.

Workflows

Workflow run notifications

Receive events when workflows start, progress, or finish.

Sync

Event-driven sync

Trigger your own jobs when webhook events arrive.

Chaining

Integration chaining

Forward webhook events into your message bus, ETL, or internal services.

Custom Code

Run custom code actions

Run your own code as workflow steps inside the workflow engine when you need custom handling.

Workflow step execution

Run your custom code as an action inside the workflow engine.

Custom integration logic

Implement connector-specific handling that doesn't fit standard connectors.

Payload shaping

Transform inputs and outputs to match downstream APIs and systems.

Controlled extensions

Add targeted logic without moving workflow execution outside Next Gate Tech.

From the Engineering Blog

View All
Beyond the Agent Harness
·Engineering & AI

Beyond the Agent Harness

As agents become easier to build, the real challenge is governing them through a shared control layer that keeps permissions, policies, approvals, ownership, and auditability consistent across the entire operation.

Building Agents We Can Measure
·Engineering & AI

Building Agents We Can Measure

Why we build Spark with evaluation-driven development, using tasks, trials, traces, and evaluators to make agent behavior visible from design to production.

Build With Agents. Run Without Them.
·Product

Build With Agents. Run Without Them.

Why agentic governance and deterministic execution are the disciplines that separate a compelling demo from a system you can actually run in production.

Start Building

Explore our API documentation and start integrating today

View API / MCP DocsContact us