Key Concepts
- Native MCP Node (NAN Team): MCP server trigger released by the NAN team, allowing interaction with tools and workflows from third-party apps.
- Community MCP Node: MCP node created by community members, enabling connection to various third-party MCP servers.
- MCP (Model Context Protocol): A universal plug for AI agents, simplifying workflow interactions.
- VPS (Virtual Private Server): A virtualized server that provides a local hosting environment accessible from anywhere.
- Cloud Hosting: Hosting Naden on a cloud platform, typically more expensive than using a VPS.
Native MCP Node (NAN Team)
- Functionality: The native MCP node, released by the NAN team, acts as a trigger that allows you to attach multiple tools and workflows. It enables interaction with these tools directly from third-party applications like a cloud desktop app.
- Example: The video references a previous tutorial demonstrating the use of the native MCP server. This server has a URL that can be accessed, and tools like Pinecone vector databases or even entire separate workflows can be attached.
- Interaction: The AI agent can utilize the MCP client node to call different tools attached to the MCP server. Instead of specifying which tool to use, the AI agent interacts with the MCP client node, which then manages the tool selection.
- Advantages:
- Enables interaction with the NAD workflow from third-party applications.
- Limitations:
- Currently, it cannot interact with third-party MCP servers available through the community node.
Community MCP Node
- Requirement: Requires a locally hosted Naden instance or a VPS (Virtual Private Server) because it doesn't work on the native cloud account.
- Installation: The video promotes Hostinger as an easy way to install Naden on a VPS. The KVM2 plan is recommended.
- Step 1: Click the provided link to access Hostinger.
- Step 2: Choose the KVM2 plan.
- Step 3: Select a billing period (12 or 24 months recommended for best deal).
- Step 4: Apply the "AI workshop" coupon code for an additional 10% discount.
- Step 5: Search for "Naden" in the Hostinger dashboard.
- Step 6: Click on Naden and confirm the installation.
- Step 7: Complete the payment and access the Naden VPS through the Hostinger dashboard.
- Advantages:
- Access to Third-Party MCP Servers: The primary advantage is the ability to connect to a wide range of third-party MCP servers offered by various companies (e.g., Appify, Alibaba, AWS, Box).
- Universal Plug: The MCP acts as a universal plug for AI agents, simplifying interactions with different services.
- Limitations:
- Currently, there is no known way to use the community node with a cloud desktop app or other third-party applications.
Comparison and Use Cases
- MCP Servers: The video highlights the growing number of MCP servers available, emphasizing their potential to streamline AI agent workflows.
- Model Context Protocol (MCP): Defined as a "universal plug" for AI agents, MCP simplifies interactions and management of workflows.
- Trade-offs: The choice between the native and community MCP nodes depends on the specific use case. If interaction with third-party MCP servers is crucial, the community node is preferred. If integration with a cloud desktop app is essential, the native node might be more suitable, pending future updates.
- Quote: "Essentially the MCP or a model context protocol is like a universal plug for your AI agents." This emphasizes the role of MCP in simplifying AI agent interactions.
Conclusion
The video provides a detailed comparison of the native and community MCP nodes for Naden. The native node offers integration with third-party apps but lacks connectivity to external MCP servers. The community node, while requiring local hosting or a VPS, provides access to a vast ecosystem of third-party MCP servers. The choice between the two depends on the specific needs of the user, with the community node currently favored due to its broader compatibility with external services. The video expresses hope that the NAN team will update the native MCP node to include support for third-party MCP servers in the future.
AI summaries can miss context or contain errors. Check important details against the original video.





