> For the complete documentation index, see [llms.txt](https://docs.scaleconnect.com.au/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.scaleconnect.com.au/scaleconnect-docs/developer-docs/mcp.md).
# MCP
The Model Context Protocol (MCP) is an open standard that lets AI tools connect to external data sources and take actions on your behalf. ScaleConnect supports MCP-compatible integrations, meaning you can connect AI assistants like Claude, ChatGPT, Cursor, and other tools directly to your ScaleConnect data — giving them the ability to read client portfolios, run simulations, pull dashboard metrics, and more.
***
## What Is MCP?
MCP is a protocol that standardises how AI tools communicate with external systems. Instead of each AI tool building a custom integration for every data source, MCP provides a common interface. An AI assistant that supports MCP can connect to any MCP-compatible data source using the same pattern.
Think of it as a universal adapter. Your AI assistant speaks MCP, ScaleConnect's API responds to MCP-formatted requests, and the result is a seamless connection where the AI can access your broker data as naturally as it reads a document or searches the web.
The protocol is gaining broad adoption across AI platforms:
| Platform | MCP Support |
| ---------------------- | ---------------------------------------------------------------- |
| **Claude (Anthropic)** | Native MCP support via tool use |
| **ChatGPT (OpenAI)** | MCP support via plugins and function calling |
| **Cursor** | Built-in MCP server configuration |
| **Custom AI agents** | Any agent framework that implements the MCP client specification |
***
## How ScaleConnect Supports MCP
ScaleConnect's [REST API](/scaleconnect-docs/developer-docs/open-api.md) provides the data layer that MCP-compatible tools consume. Every endpoint documented in the API is accessible to an AI tool configured with the right credentials and MCP client adapter.
### Current architecture
Today, the integration works through a REST API with MCP client adapters. You configure your AI tool's MCP client to point at the ScaleConnect API base URL, provide a Bearer token for authentication, and map the API endpoints to MCP tool definitions. The AI tool then treats each endpoint as a callable function.
This approach works with any AI platform that supports MCP or function calling, and gives you full control over which endpoints the AI can access.
### Dedicated MCP server (roadmap)
A dedicated MCP server binary is on the ScaleConnect roadmap. This will provide a turnkey MCP integration — install the server, provide your credentials, and your AI tool automatically discovers all available ScaleConnect functions without manual endpoint mapping. Until then, the REST API approach provides the same capability with some additional setup.
{% hint style="info" %}
The REST API approach is fully functional today. The dedicated MCP server will simplify setup but won't add new capabilities — the same data and actions will be available either way.
{% endhint %}
***
## What AI Tools Can Access
Once connected, your AI tool can interact with the full breadth of ScaleConnect data. The available capabilities map directly to the [API endpoint groups](/scaleconnect-docs/developer-docs/open-api.md):
| Capability | What the AI can do |
| --------------------------- | ------------------------------------------------------------------------------------------------------------------------------ |
| **Client portfolio data** | Read client lists, property details, loan structures, equity positions, LVR, and cashflow data |
| **Simulation engine** | Create and run strategies — add events, model scenarios, project outcomes |
| **Dashboard metrics** | Pull Loan Book LTV, commission breakdowns, retention rates, and pipeline status |
| **Calendar and scheduling** | Check availability, create events, book meetings through [Google Calendar](/scaleconnect-docs/integrations/google-calendar.md) |
| **Agent sessions** | Create or continue AI conversation sessions — useful for chaining ScaleAI interactions |
| **Review data** | Access AI Review history, pending reviews, and review content |
| **Goals and progress** | Read and update client goals, track progress against targets |
| **Charts and projections** | Retrieve the data behind ScaleConnect's charts — equity growth, cashflow projections, net worth over time |
***
## Use Cases
Connecting AI tools to ScaleConnect data opens up workflows that go beyond what either system can do alone.
### Custom AI workflows
Build AI agents that perform multi-step analysis across your client book. For example, an agent that:
1. Pulls all clients in the Month readiness tier from the [Client Readiness System](https://github.com/scaleapp-au/scaleconnect-docs/blob/main/integrations/client-readiness.md)
2. Runs a refinance simulation for each one using current market rates
3. Compares the projected savings against their existing loan
4. Drafts a personalised outreach message for each client who would benefit
This kind of workflow combines ScaleConnect data with AI reasoning to produce actionable output at scale.
### Automated reporting
Connect ScaleConnect to an AI tool that generates weekly or monthly business reports. The AI pulls dashboard metrics, commission data, and pipeline status through the API, then formats the data into a narrative report with insights and recommendations. Deliver it to your inbox every Monday morning without manual effort.
### Smart scheduling
An AI assistant with access to both your ScaleConnect data and Google Calendar can intelligently manage your meeting schedule. It identifies which clients are ready for a review, checks your availability, proposes optimal meeting times, and books the meeting — all in a single conversational exchange.
### Portfolio analysis
Ask your AI assistant questions about your client book in natural language:
* "Which of my clients have LVRs above 60% and are paying above-market rates?"
* "What's the total trail commission I'd lose if my five least-engaged clients left?"
* "Show me clients whose available equity has grown by more than $50,000 in the last 6 months"
The AI translates these questions into API calls, retrieves the data, and presents the answer in plain language.
### CRM enrichment
If you use a CRM alongside ScaleConnect, an AI agent can keep both systems in sync. It reads client updates from ScaleConnect, identifies changes (new property, loan settled, goal updated), and pushes the relevant data to your CRM — maintaining a single source of truth without manual data entry.
***
## Getting Started
Setting up MCP access to ScaleConnect requires three steps: API credentials, client configuration, and endpoint mapping.
### Step 1: Obtain API access
You need a valid Bearer token to authenticate API requests. See [Open API Overview](/scaleconnect-docs/developer-docs/open-api.md) for details on authentication and token management. Store your token securely — it provides full access to your ScaleConnect data.
### Step 2: Configure your MCP client
Each AI platform has its own way of configuring MCP connections. The general pattern is:
1. Open your AI tool's MCP or function-calling configuration
2. Set the base URL to your ScaleConnect API endpoint
3. Add your Bearer token as the authentication credential
4. Define the available tools (endpoints) the AI can call
For Cursor, this means adding an entry to your MCP server configuration file. For Claude, you configure tools through the API or desktop app settings. For custom agents, implement the MCP client specification in your agent framework.
### Step 3: Map endpoints to tools
Define which ScaleConnect API endpoints your AI tool can access. Each endpoint becomes a "tool" or "function" that the AI can call. At minimum, you'll want to map:
| Tool | Endpoint group | Purpose |
| --------------------- | ---------------- | ------------------------------------------------------------------------ |
| **List clients** | Connect | Retrieve your client list with readiness and LVR data |
| **Get client detail** | User Profile | Pull full profile, properties, loans, and cashflow for a specific client |
| **Run simulation** | Simulation | Create and execute strategy scenarios |
| **Get dashboard** | Connect / Charts | Pull LTV, commission, and pipeline metrics |
| **Check calendar** | Google Calendar | Query availability and book meetings |
Start with a small set of tools and expand as your workflows develop. The Swagger documentation at `/api` provides the full schema for each endpoint, which you can use to define tool parameters and response types.
***
## Security Considerations
When connecting AI tools to ScaleConnect data, keep these practices in mind:
* **Token scope** — your Bearer token provides full access to your account. Only share it with tools you trust.
* **Data sensitivity** — client financial data is sensitive. Ensure your AI tool's data handling policies align with your obligations under privacy legislation and your client agreements.
* **Audit trail** — API requests made by AI tools appear in the same logs as manual requests. You can review what was accessed and when.
* **Principle of least privilege** — map only the endpoints your AI tool needs. If it only analyses portfolios, don't give it access to scheduling or messaging endpoints.
For more on data handling and privacy, see [Data & Privacy](https://github.com/scaleapp-au/scaleconnect-docs/blob/main/integrations/data-and-privacy.md).
***
Do I need to be a developer to set this up?
The current REST API approach requires some technical knowledge — configuring API endpoints, managing tokens, and setting up MCP clients. If you're not comfortable with this, ask your IT team or a developer to help with the initial setup. Once configured, the day-to-day use is conversational.
Will a dedicated MCP server simplify this?
Yes. When the dedicated MCP server is released, setup will be as simple as installing the server and providing your credentials. The server will automatically expose all available ScaleConnect functions to any MCP-compatible AI tool, eliminating the need for manual endpoint mapping.
Can I restrict which data the AI sees?
Currently, the Bearer token provides access to all data associated with your account. You can control what the AI accesses by limiting which endpoints you map as tools. Granular permission scopes (e.g., read-only, specific client segments) are on the roadmap.
Is MCP the same as the ScaleAI features built into ScaleConnect?
No. ScaleAI features — [Ask AI](https://github.com/scaleapp-au/scaleconnect-docs/blob/main/integrations/product-docs/scaleconnect-docs/scaleai/ask-ai.md), [AI Reviews](https://github.com/scaleapp-au/scaleconnect-docs/blob/main/integrations/ai-reviews.md), [Broker Brain](https://github.com/scaleapp-au/scaleconnect-docs/blob/main/integrations/broker-brain.md) — are built-in AI capabilities that run inside ScaleConnect. MCP is a protocol for connecting *external* AI tools to ScaleConnect data. They complement each other: use ScaleAI for in-platform intelligence and MCP for extending that intelligence to your own tools and workflows.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.
Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:
```
GET https://docs.scaleconnect.com.au/scaleconnect-docs/developer-docs/mcp.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.