> For the complete documentation index, see [llms.txt](https://docs.caudena.com/VWRJyWGSviHEMRPlWDXF/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.caudena.com/VWRJyWGSviHEMRPlWDXF/model-context-protocol-mcp.md).

# Model Context Protocol (MCP)

## Overview

Prism MCP Server exposes Clain's blockchain intelligence capabilities through the Model Context Protocol, allowing AI assistants (like Claude Desktop, Cursor, etc.) to directly interact with Prism's data.

This integration enables LLMs to:

* **Investigate** addresses, transactions, and clusters.
* **Trace** fund flows across multiple blockchains (UTXO and EVM).
* **Analyze** risk exposure and entity attribution.
* **Detect** cross-chain swaps and bridges.

{% hint style="info" %}
All tools respect Prism's **inverse risk scoring** system (1.0 = Severe Risk, 10.0 = Lowest Risk) and entity categorization.
{% endhint %}

## Installation

When installing the MCP for various AI tools, it's important to note that there may be specific requirements or steps unique to each software. For a comprehensive understanding, refer to the user documentation provided by the specific vendor. This guide outlines the installation process for popular tools such as Cursor, Claude, and ChatGPT, ensuring you have the necessary information to integrate Prism MCP server successfully with these applications.

### Cursor

You can add the Prism MCP server to Cursor following the next steps:

* Click the **gear wheel (Settings)** icon (top-right).
* In the left sidebar, open **Tools & MCP**.
* Click **New MCP Server**.
* Add the snippet below to mcp.json:

```json
{
  "mcpServers": {
    "prism": {
      "url": "https://prism-api.caudena.com/mcp"
    }
  }
}
```

* Save to enable it.
* Sanity check: it should appear under **Installed MCP Servers** and show tools enabled when connected.

<figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FWglCcuHyMAwasXZ82J0R%2Fmcp-cursor-1.png?alt=media&amp;token=66444e94-c8e3-47f4-bbdd-99422e729cc2" alt=""><figcaption></figcaption></figure>

<figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FNzrKU3r9JSZMjpJB2boD%2Fmcp-cursor-2.png?alt=media&amp;token=3e9c374f-0d5b-4807-ab45-2fdef05b912e" alt=""><figcaption></figcaption></figure>

### Claude.ai

To add the Prism MCP to Claude:

1. Open **Settings → Connectors**
2. Click **Add custom connector**.
3. Fill in:
   * **Name**: `Prism`
   * **URL**: `https://prism-api.caudena.com/mcp`
4. Click **Add**.
5. **MCP** will request access to your **Caudena Profile** - **Allow Access.**

**Sanity check:** Prism should appear in your connectors list and be connectable/usable in chats.

<div><figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FZfj6UpzAmOAnZxiPosGP%2Fmcp-claude-1.png?alt=media&amp;token=f7dfac0e-7f7d-4588-9a81-9a889ab9b28a" alt=""><figcaption></figcaption></figure> <figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FyNob79DlBhaUMcdAamhl%2Fmcp-claude-2.png?alt=media&amp;token=fa3f310d-1d8a-4edf-a2e5-5990d157cce2" alt=""><figcaption></figcaption></figure></div>

### ChatGPT

1. Open **Settings → Apps → Advanced settings**
2. Turn on **Developer mode (BETA)**.
3. Click **Create app**.
4. In **New App**:
   * **Name**: `Prism`
   * **Description** (optional)
   * **MCP Server URL**: `https://prism-api.caudena.com/mcp`
   * **Authentication**: choose the correct option from the dropdown
     * Keep **OAuth**
5. Tick the risk acknowledgement checkbox.
6. Click **Create**
7. **MCP** will request access to your **Caudena Profile** - **Allow Access.**

**Sanity check:** the app should appear under enabled apps and be available to use in chats.

<div><figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FQ0I4KqIak8yGRTVmAMXI%2Fmcp-chatgpt-1.png?alt=media&amp;token=3814f8b0-f913-4769-be7d-7a11ebbe0866" alt=""><figcaption></figcaption></figure> <figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FLLXXI4YuF3M4XaFbUP1l%2Fmcp-chatgpt-2.png?alt=media&amp;token=f574a1c3-7709-4980-852d-141120f75862" alt=""><figcaption></figcaption></figure></div>

<figure><img src="https://1271921339-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjrkTAYRoF1xL7YxnNjt3%2Fuploads%2FoTeNVefVsbefodj5aMAc%2Fmcp-chatgpt-3.png?alt=media&amp;token=ae9bdbaa-eb7c-48db-9519-2ca9f8c3dc9b" alt=""><figcaption></figcaption></figure>

## Capabilities

The MCP server provides three main types of capabilities:

1. **Tools**: Executable functions to query blockchain data (e.g., `prism_address_stats`, `prism_transaction_get`).
2. **Resources**: Static or dynamic context data (e.g., `prism://entity_categories`, `prism://risk_scoring_guide`).
3. **Prompts**: Pre-defined investigation workflows (e.g., `prism_address_comprehensive_analysis`).

## Supported Chains

The MCP server supports querying across all chains indexed by Prism, including:

* **UTXO**: Bitcoin (BTC), Litecoin (LTC), Dogecoin (DOGE)
* **EVM**: Ethereum (ETH), Binance Smart Chain (BNB), Tron (TRX), Base (BASE)

Use the `prism://supported_chains` resource to get a live list of supported currencies.


---

# 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.caudena.com/VWRJyWGSviHEMRPlWDXF/model-context-protocol-mcp.md?ask=<question>&goal=<user_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.