Use Rodz with Claude Desktop via MCP: AI-Powered Business Intelligence

In a nutshell: The Model Context Protocol (MCP) lets Claude Desktop call external APIs directly. By connecting Rodz as an MCP server, you can query intent signals, enrich company data and analyze prospects through plain English conversation. This guide walks you through installing Claude Desktop, configuring the Rodz MCP server, and running your first AI-powered business intelligence queries. Setup takes about fifteen minutes.

What Is MCP and Why Does It Matter for B2B Sales?

The Model Context Protocol, or MCP, is an open standard created by Anthropic that lets AI assistants like Claude connect directly to external tools, databases and APIs. Think of it as a universal adapter: instead of copying data into a chat window, you give Claude permission to query a service on your behalf, in real time, using structured API calls behind the scenes.

Before MCP, using AI with your business data meant a lot of manual work. You would export a CSV from your CRM, paste it into a prompt, wait for a response, then manually act on the results. MCP eliminates that friction. Claude can call the Rodz API, retrieve the exact signals or enrichment data you need, and reason about the results in a single conversation.

For B2B sales teams, this changes the game. Instead of writing cURL commands or building custom dashboards, you can ask Claude questions like “Which of my target accounts raised funding in the last 30 days?” or “Enrich this list of domains and find the decision-makers.” Claude handles the API calls, parses the JSON, and gives you a structured answer you can act on immediately.

MCP is not a proprietary Anthropic feature locked to one product. It is an open protocol with a published specification, which means any tool or platform can implement it. Rodz provides an MCP server that speaks this protocol, making the entire Rodz API accessible from Claude Desktop without writing a single line of integration code.

If you want to understand the Rodz API endpoints that power the MCP integration, the Rodz API Reference covers every endpoint, rate limit and error code in detail.

Prerequisites

Before you begin, make sure you have the following:

A Rodz account with API access. You need an active API key from your Rodz dashboard. If you have not created one yet, follow the authentication and first request guide to get your key.
Claude Desktop installed. Claude Desktop is Anthropic’s native application for macOS and Windows. You will install it in Step 1 below if you do not have it already.
Node.js 18 or later. The Rodz MCP server runs as a Node.js process. Download Node.js from https://nodejs.org if you do not have it. Verify your installation by running node --version in your terminal.
A text editor. You will need to edit a JSON configuration file. VS Code, Notepad++, or any editor that handles JSON will work.
Basic comfort with the command line. You will run a few terminal commands during setup, but nothing complex. If you can navigate directories and run npm install, you are set.
Your Rodz API key on hand. Copy it from Settings > API in your Rodz dashboard. You will paste it into the MCP configuration file during setup.

Step 1: Install Claude Desktop

If you already have Claude Desktop installed and running, skip ahead to Step 2.

Download the application

Visit https://claude.ai/download and download the installer for your operating system. Claude Desktop is available for macOS (Apple Silicon and Intel) and Windows (64-bit).

Run the installer

On macOS, open the downloaded .dmg file and drag Claude to your Applications folder. On Windows, run the .exe installer and follow the prompts. The default installation options are fine for most users.

Launch Claude Desktop and sign in with your Anthropic account. If you do not have one, you can create a free account during the sign-in flow. Once you are signed in, you will see the main chat interface.

Verify MCP support

MCP support is built into Claude Desktop and requires no additional plugins. To confirm your version supports MCP, open Settings (the gear icon in the bottom-left corner) and look for the Developer section. You should see a menu item for Edit Config or a reference to MCP servers. If you see it, you are ready to proceed.

Step 2: Install the Rodz MCP Server

The Rodz MCP server is a lightweight Node.js package that translates Claude’s MCP requests into Rodz API calls. It handles authentication, request formatting, and response parsing so Claude can interact with the full Rodz API.

Install the package globally

Open your terminal and run:

npm install -g @rodz/mcp-server

This installs the Rodz MCP server as a global command-line tool. After installation, verify it is available:

rodz-mcp --version

You should see a version number printed to the console.

Alternative: use npx without installing

If you prefer not to install the package globally, you can use npx to run the server on demand. Claude Desktop supports this approach natively. You will see how to configure it in the next step.

Step 3: Configure MCP in Claude Desktop

Claude Desktop uses a JSON configuration file to know which MCP servers are available. You need to add the Rodz server to this file.

Locate the configuration file

The configuration file location depends on your operating system:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

If the file does not exist, create it. If it already exists (perhaps with other MCP servers configured), you will add the Rodz entry alongside the existing ones.

Add the Rodz MCP server

Open the configuration file in your text editor and add (or merge) the following JSON structure:

{
  "mcpServers": {
    "rodz": {
      "command": "rodz-mcp",
      "args": ["--stdio"],
      "env": {
        "RODZ_API_KEY": "your_api_key_here"
      }
    }
  }
}

Replace your_api_key_here with your actual Rodz API key.

If you prefer to use npx instead of a global install, replace the command and args fields:

{
  "mcpServers": {
    "rodz": {
      "command": "npx",
      "args": ["-y", "@rodz/mcp-server", "--stdio"],
      "env": {
        "RODZ_API_KEY": "your_api_key_here"
      }
    }
  }
}

Understanding the configuration

Each field serves a specific purpose:

command: The executable Claude Desktop will launch. Either the globally installed rodz-mcp binary or npx for on-demand execution.
args: Command-line arguments passed to the server. The --stdio flag tells the server to communicate through standard input/output, which is the transport MCP uses with Claude Desktop.
env: Environment variables passed to the server process. Your API key is provided here so it never needs to be typed into the chat interface.

Restart Claude Desktop

After saving the configuration file, quit Claude Desktop completely and relaunch it. On macOS, make sure to use Quit from the menu bar (not just close the window). On Windows, close the application from the system tray.

Verify the connection

When Claude Desktop restarts, look for a small hammer icon or a tools indicator near the chat input field. Click it to see the list of available MCP tools. You should see Rodz-related tools such as search_signals, get_signal, enrich_company, and enrich_contact. If the tools appear, the connection is working.

If you do not see the tools, check the MCP logs. On macOS, logs are at ~/Library/Logs/Claude/mcp*.log. On Windows, check %APPDATA%\Claude\logs\. Common issues include an incorrect API key, a missing Node.js installation, or a typo in the JSON config file.

Step 4: Your First MCP Conversation

Now that Rodz is connected, let us walk through a few practical conversations that demonstrate what you can do.

Query recent intent signals

Start a new conversation in Claude Desktop and type:

Show me all funding signals from the last 7 days for companies in the SaaS industry.

Claude will use the Rodz MCP server to call the signals endpoint, filter by signal type and date range, and return a structured summary. You might see something like:

I found 23 funding signals in the last 7 days for SaaS companies. Here are the highlights:

Acme Corp - Series B, $45M raised on March 8, 2026. Lead investor: Sequoia Capital.

DataFlow Inc - Seed round, $3.2M raised on March 7, 2026. Lead investor: Y Combinator.

CloudMetrics - Series A, $18M raised on March 6, 2026. Lead investor: a16z. …

You can then follow up naturally: “Which of those are based in France?” or “Show me the decision-makers at Acme Corp.” Claude maintains context across the conversation, so each follow-up query builds on the previous results.

Enrich a company

Try asking:

Enrich the company with domain lemlist.com. I want to know their tech stack, employee count, and latest signals.

Claude will call the company enrichment endpoint and return structured data including firmographic details, technology stack, headcount information, and any recent signals associated with the company. This is the same data you would get from the API, but presented in a format you can immediately act on.

Find and verify contacts

You can also use Claude to find specific people at a target account:

Find the VP of Sales or Head of Sales at DataFlow Inc. Give me their email and LinkedIn profile.

Claude uses the contact enrichment endpoint to search for matching contacts, verify email addresses, and return the results in a clean format. For more on what the enrichment endpoints offer, see the Rodz API Reference.

Build a prospecting list

One of the most powerful use cases is building targeted prospecting lists through conversation:

I sell cybersecurity solutions to mid-market SaaS companies. Find me 10 companies that recently raised a Series A or Series B, have between 50 and 500 employees, and are based in Europe. For each company, find the CISO or Head of IT and give me their contact details.

This single prompt triggers multiple API calls: signal search, company enrichment, and contact enrichment. Claude orchestrates all of them and presents a ready-to-use prospecting list.

Analyze competitive movements

Intent signals go beyond funding. You can ask Claude to surface competitive intelligence:

Have any of my competitors' customers posted job offers for roles related to our product category in the last 30 days?

Claude can cross-reference signal data to identify accounts that may be evaluating new solutions, giving you timely outreach opportunities.

Step 5: Advanced Configuration and Tips

Once you are comfortable with the basics, here are a few ways to get more out of the integration.

Use system prompts for consistent output

Claude Desktop lets you set a system prompt (sometimes called “custom instructions”) that applies to every conversation. You can use this to define your ICP, preferred output format, or standard follow-up actions:

You are a B2B sales intelligence assistant. My ICP is mid-market SaaS companies
(50-500 employees) in Europe. When I ask about signals, always include the company
name, domain, signal type, date, and a one-sentence summary. Format results as a
markdown table.

With this system prompt active, every signal query will return consistently formatted tables without you having to specify the format each time.

Chain multiple queries in one conversation

MCP connections persist throughout a conversation. This means you can build complex analyses step by step:

“Show me companies that changed their CTO in the last 60 days.”
“Filter that list to companies using AWS and with more than 100 employees.”
“For the top 5, find the new CTO’s email and draft a personalized outreach message.”

Each step uses the context from the previous one. Claude remembers the results and refines the analysis progressively.

Rate limit awareness

The Rodz API has a rate limit of 100 requests per minute. When Claude runs complex queries that require multiple API calls (like enriching a list of 20 companies), it respects these limits automatically. If you hit the rate limit, Claude will let you know and suggest waiting before retrying.

For detailed rate limit information and best practices, consult the Rodz API Reference.

Export results

Claude Desktop supports copying conversation content as markdown. After building a prospecting list or signal report, you can copy the output and paste it directly into a spreadsheet, CRM import file, or Notion page. Select the relevant response, right-click, and choose Copy as Markdown.

Security Considerations

When connecting any API to an AI assistant, security matters. Here are the key points for the Rodz MCP integration:

Your API key stays local. The key is stored in the Claude Desktop configuration file on your machine. It is passed to the MCP server process as an environment variable and is never sent to Anthropic’s servers or included in conversation data.

Data flows through two services. Your queries go from Claude Desktop to Anthropic’s API (for language understanding) and from the MCP server to the Rodz API (for data retrieval). The business data returned by Rodz is processed in the conversation context. Review Anthropic’s data usage policies and your Rodz plan’s terms to ensure compliance with your organization’s data handling requirements.

Rotate your key if compromised. If you suspect your API key has been exposed, generate a new one immediately from the Rodz dashboard under Settings > API. The old key will be invalidated.

Use environment variables in shared setups. If multiple team members share a machine or configuration, consider referencing the API key from a system environment variable instead of hardcoding it in the JSON file. The MCP configuration supports environment variable expansion.

Frequently Asked Questions

What is the Model Context Protocol (MCP)?

MCP is an open protocol developed by Anthropic that standardizes how AI applications connect to external data sources and tools. It defines a client-server architecture where the AI assistant (the client) sends structured requests to an MCP server, which translates those requests into API calls, database queries, or other operations. The server returns structured results that the AI can reason about and present to the user. MCP supports multiple transport mechanisms, including standard input/output (used by Claude Desktop) and HTTP with Server-Sent Events. The protocol specification is publicly available, and any developer can build an MCP server for their tool or service.

Do I need to know how to code to use this integration?

No. The entire setup involves installing an application, running one terminal command, and editing a JSON file. Once configured, you interact with Rodz through natural language in Claude Desktop. You never need to write code, construct API requests, or parse JSON responses. Claude handles all of that through the MCP connection. That said, if you are a developer, you can also use the Rodz API directly for more advanced integrations. The Getting Started guide covers that path.

Which Claude plan do I need for MCP?

MCP is available on all Claude plans that include Claude Desktop, including the Pro plan. The free tier of Claude may have limitations on the number of messages per day, but MCP functionality itself is not restricted by plan. Check Anthropic’s current pricing page for the latest details on message limits and plan features.

What Rodz API endpoints are available through MCP?

The Rodz MCP server exposes the full Rodz API surface. This includes signal search and retrieval, company enrichment, contact enrichment, webhook management, and signal configuration. Every endpoint documented in the Rodz API Reference is accessible through MCP. Claude translates your natural language queries into the appropriate endpoint calls, selects the right parameters, and combines results from multiple endpoints when needed.

How does MCP handle errors and rate limits?

When an API call fails or hits a rate limit, the MCP server returns the error information to Claude, which then explains the issue in plain language. For example, if you exceed the 100-requests-per-minute rate limit, Claude will tell you that the API is temporarily throttled and suggest waiting before retrying. Authentication errors, invalid parameters, and server errors are all handled gracefully. You never see raw error codes unless you specifically ask for the technical details.

Can I use other MCP servers alongside Rodz?

Yes. Claude Desktop supports multiple MCP servers simultaneously. You can add as many servers as you need to the mcpServers object in your configuration file. For example, you could connect Rodz for intent signals, a CRM MCP server for deal data, and a Slack MCP server for messaging, all in the same Claude Desktop instance. Each server operates independently, and Claude can use tools from any of them within a single conversation.

Is the MCP connection real-time?

The MCP server makes live API calls to Rodz every time Claude needs data. There is no caching layer between Claude and the Rodz API. This means you always get the most current signals and enrichment data available. The latency is typically low: a signal search query completes in one to three seconds, and enrichment calls take two to five seconds depending on the depth of data requested.

Can I restrict which endpoints Claude can access?

The Rodz MCP server exposes all available endpoints by default. If you need to restrict access, for example limiting Claude to read-only signal queries and blocking webhook management, you can configure this at the API key level in the Rodz dashboard. Create a dedicated API key with restricted permissions and use that key in your MCP configuration. This way, Claude can only access the endpoints your restricted key allows, regardless of what the MCP server supports.

What to Do Next

You now have Claude Desktop connected to the Rodz API through MCP. Here are a few logical next steps:

Explore the full API documentation at https://api.rodz.io/docs to understand every signal type and enrichment field available. The more you know about what data Rodz provides, the better questions you can ask Claude.
Set up a daily signal review. Start each morning by asking Claude for new signals matching your ICP. This takes thirty seconds and replaces a manual dashboard check.
Build reusable prompts. Save your best-performing queries as text snippets. A prompt like “Find European SaaS companies that raised funding this week, enrich them, and find the CTO’s email” can become a daily workflow.
Combine with webhooks. For fully automated workflows, set up Rodz webhooks to push signals to your CRM or Slack, and use Claude Desktop for ad-hoc analysis and deep dives. The two approaches complement each other well.
Share the setup with your team. The configuration file and installation steps are simple enough for any team member to replicate. Standardize your team’s access to Rodz intelligence through Claude Desktop.

The combination of structured intent signal data and natural language AI creates a workflow that is faster and more flexible than either tool alone. You get the precision of an API with the accessibility of a conversation.