N8n AI Agent DVM MCP Client

🚀 n8n AI Agent for DVM MCP

This project creates an AI agent that serves as a user of the Data Vending Machine (DVM) for the Model Context Protocol (MCP) server tool running over the Nostr network.

When the agent receives a prompt requiring the use of tools:

• It will query the network to find available tools it can use.
• Send requests to these tools.
• Wait for the required tool responses.
• Read the tool responses and reply to the user.

Why is this interesting?

• For a Large Language Model (LLM) to use MCP tools, the MCP server usually needs to be installed on the same system.
• There has been no way for an LLM to discover and use MCP servers and tools that are not yet installed.
• The Data Vending Machine enables the LLM to discover and use tools via the Nostr network.
• This AI agent example demonstrates how to build an AI that can find and use MCP tools it hasn't installed by querying DVMs on the Nostr network.

🔧 Technical Details

• MCP: The Model Context Protocol is an open protocol that allows AI agents to access data sources and tools. Learn more.
• DVM: The Data Vending Machine listens for and responds to queries over the Nostr network. Learn more.
• DVMCP: The Data Vending Machine Context Protocol connects the Model Context Protocol server to the Nostr Data Vending Machine ecosystem. Any MCP server can offer any tool over the network, and any agent client can access all tools provided by any server. See DVMCP developed by Gzuuus and NostrDVM developed by BelieveTheHype.
• n8n: A low - code platform well - suited for creating Langchain tool agents. Learn more.

Nostr Link

• Contact me npub1r0d8u8mnj6769500nypnm28a9hpk9qg8jr0ehe30tygr3wuhcnvs4rfsft.

📄 License

• This project is licensed under the MIT License.

🚀 Quick Start

Installation and Configuration Guide

1. Install n8n

Before you start, make sure n8n is installed. You can find detailed installation instructions in the n8n official documentation.

2. Configure Credentials

Configure the corresponding credentials (such as API keys, access tokens, etc.) according to your use case. This information will be used to interact with the MCP server and DVM.

3. Set System Prompt

Set a system prompt for the AI agent, for example:

**Tips for Using DVM**

- If the user requests a tool function you don't have, query the MCP server tools on the Nostr network. After finding the relevant tool, ask the user if they want to use it.

- If the user requests to use a certain tool, follow these steps:

1) Send a request to the MCP server.
2) Wait for 15 seconds.
3) Read the MCP server's response.
4) If no response is received from the MCP server within 45 seconds, repeat steps 2 and 3.
5) Complete any final steps or required localization (e.g., query the exchange rate and convert US dollars to the local currency).

4. Configure Workflow

Configure the n8n workflow according to your specific needs. You can refer to the n8n documentation to learn how to create and edit workflows.

5. Test the Agent

Before actual use, conduct thorough testing to ensure that the agent can correctly interact with the MCP server and DVM and call tools when needed.

💻 Usage Examples

Financial Data Analysis

The user requests real - time stock prices. The AI agent will obtain data by querying relevant DVMs on the Nostr network and convert it to the user's local currency.

Logistics Optimization

The user requests to plan the optimal transportation route. The AI agent will use its built - in path - planning tool or call an external service by querying the DVM to complete the task.

🐞 Error Handling and Debugging

In practical applications, you may encounter the following issues:

1. Network Connection Issues: Ensure that your n8n instance can access the Internet and that the Nostr nodes are running normally.
2. MCP Server Unavailable: If the target MCP server is offline or unresponsive, try other available servers or tools.
3. Credential Errors: Check that the configured credentials are correct.

📚 Documentation

• n8n official documentation
• Model Context Protocol (MCP)
• Nostr network documentation
• DVMCP development documentation