You can now connect to dozens of MCP servers directly from the Node panel with a simple OAuth flow. Pick a server, sign in, and your agent is instantly ready to work. The team rolled out a fresh batch of MCP connections right on the panel, so you avoid hunting for separate integrations. New additions include Airtable, Grafana, New Relic, Jotform, and PandaDoc. They join existing options like Notion, Stripe, GitLab, Apify, Linear, monday.com, and Hugging Face. Connecting to nearly 70 services is now a quick OAuth flow.
What exactly changed, and how does it work now?
The change is straightforward: you connect dozens of MCP servers directly from the Node panel via OAuth. You pick a server, complete sign‑in, and tools appear for the agent immediately.
The team just shipped a new batch of servers available from the panel. The list now includes Airtable, Grafana, New Relic, Jotform, and PandaDoc. These join already available connections like Notion, Stripe, GitLab, Apify, Linear, monday.com, and Hugging Face.
Connecting to nearly 70 services is now a quick OAuth flow. You skip manual integrations and scattered setups. After authorization, the agent instantly sees the right tools, and you control what is exposed.
Connecting to nearly 70 services is now a quick OAuth flow.
Convenient, right? But what does that mean for everyday workflows and real use? Let’s look at a concrete Notion example and compare the before and after.
Why does this matter in practice if you already use these services?
The unlock is immediate if you work with these daily. The agent gains tool access without extra steps or waiting.
Take Notion. Properly wiring an agent used to mean asking an admin for permissions, creating an internal integration, sharing pages with it, and configuring multiple Notion tools for different operations. That took time and split team attention.
Now it’s just an OAuth click. The agent can pull meeting notes for context, draft a page from the conversation, and update a database when something happens elsewhere. You do not build many tiny actions — the capabilities arrive with the connection.
The same approach applies to the other MCP servers listed. Setup remains similarly simple, and you still control which tools are visible. Want to limit certain capabilities? Just toggle them off.
How do you set it up today, and what’s next?
All of these MCP server connections are similarly simple to set up. You choose a server from the Node panel, authorize, and then manage what the agent can see and use.
More connections are on the way. They start with more of the most‑used MCP servers that support OAuth and Dynamic Client Registration (DCR). That means even more quick hookups with fewer manual steps.
Need something not yet listed? Use the MCP Client Tool. It lets you point at any MCP endpoint with a bit more setup. You still get access to unlisted servers if you are ready to do a few extra steps.
Pick a server, sign in, and it's ready for your agent to use.
Ready to connect your first services? Before that, decide when to pick a node, an agent tool, or a full MCP server. That choice helps you avoid extra calls and confusion.
When should you choose a node, an agent tool, or an MCP server?
In short: nodes are for deterministic steps, agent tools are fixed actions the agent triggers, and MCP servers are broad toolsets where the agent decides what to do.
Native nodes are the right call for deterministic workflows. If you know the exact API call at a given step, or you are building something that must behave the same every time, a node gives fine‑grained control over inputs, outputs, and error handling. Use nodes wherever the logic does not require a decision.
Example: every new customer should get a welcome page in your team wiki. The Notion node creates that page in the same database, with the same fields, every single time. No judgment call is needed, so a node is all you need.
When you add a node to your AI Agent as a tool, the agent chooses when to call it, but you still define the exact action ahead of time. Because the tool is fixed and well‑defined, the agent needs less reasoning, which often means fewer LLM calls than with a broader toolset.
Example: your support agent has a Notion “Log escalation” tool. It is one specific action, pointed at one specific database. The agent decides when an issue needs escalating, but it cannot do anything in Notion beyond that job.
MCP servers shine when you are letting an agent decide what to do. Connect once, and the agent gets a whole toolset it can pick from as needed. You can still toggle individual tools on or off to decide what is exposed, but the agent decides which to use, when, and how. That flexibility comes with more reasoning on the agent’s side.
MCP servers shine when you’re letting an agent decide what to do.
Example: connect the Notion MCP server, and your agent can search pages, pull meeting notes, create a draft, or update a database entry — whatever fits the request. You did not build a tool for each action; they arrived with the connection.
You don’t have to pick just one
None of these replace the others. Nodes give you precision at a workflow step, agent tools give limited discretion over fixed actions, and MCP servers give a broad surface for the agent to reason over. Depending on the goal, you will likely mix all three.
In real workflows, you often blend approaches. A deterministic node creates a page, an agent tool decides when to escalate, and an MCP server provides context and extra actions. You control tool visibility and set the boundaries, and the agent works within them.
None of these replace the others.
Start with the most needed OAuth connections and expand as required. Use nodes for predictability, tools for timing choices, and MCP for flexibility. That balance yields the best mix of control and capability.
Based on {source_source}.
