🧩 @port-labs/n8n-nodes-portio-experimental

Port.io n8n custom nodes for interacting with Port via n8n workflows.

As of November 2025, the focus of this package is on AI interactions through Port’s AI APIs. Future releases will expand to additional REST integrations and workflow automations, enabling n8n and Port users to connect multiple tools seamlessly. This is an experimental package, no support can be provided in relation to your use of this package, if you've found a bug or need a feature currently not supported, open a GitHub issue,

---

🚀 Installation

Option 1 — Install via npm (recommended)

You can install the package directly from npm:

npm install @port-labs/n8n-nodes-portio-experimental

or using pnpm:

pnpm add @port-labs/n8n-nodes-portio-experimental

Note

When installed this way, the nodes will be available in your n8n instance once you configure n8n to load community nodes. To enable this you must set the environment variable N8N_COMMUNITY_PACKAGES_ENABLED=true and install the npm packages in N8N_HOME.

---

Option 2 — Local installation (development or custom build)

Prerequisites

• 🧱 A working n8n instance (see this quick-start guide for setup via Docker Compose)
• 📦 pnpm installed globally (npm install -g pnpm)

Steps

1. Clone this repository inside your n8n custom nodes directory:

   mkdir -p ./n8n-data/custom
   cd ./n8n-data/custom
   git clone https://github.com/port-labs/n8n-nodes-portio-experimental.git
   cd n8n-nodes-portio-experimental

2. Install dependencies and build the TypeScript source:

   pnpm install
   pnpm build

3. If the build completes successfully, you’ll see a new dist directory containing the compiled nodes.

---

🧭 Quick Start

Adding a Port node in n8n

1. Create a new workflow in your running n8n instance.
   

2. Set the trigger to “Manual Trigger” so you can execute it by clicking.
   

3. Add a new node (+ button) and search for port.
   

4. Select the Port.io node and choose “General Purpose AI Interaction”.
   

5. Connect the trigger node to the Port node.
   

🎉 You’ve just deployed your first Port n8n node!

---

🔐 Adding Your Port Account

You can store multiple Port credentials in n8n, but we’ll start with one.

Requirements

• A Port.io account with AI features enabled
• Your Client ID and Client Secret (see Port API guide)

Steps

1. Double-click your Port node.
2. Under “Credential to connect with”, choose “Create new credential.”
   
3. Enter your Port client_id and client_secret, then click Save.
   

Note

JWT authentication is currently not supported because it’s unsuitable for long-term credential storage. OAuth2 credentials are the recommended method.

---

🧠 Running the Node

Before execution, ensure:

• You have a Manual Trigger connected to a Port node
• Credentials are saved
• Your Port account has AI features enabled

Then:

1. Open your AI node and configure the parameters (for example, set a user prompt).
   

2. Execute the workflow manually and inspect the node output.
   

✅ If successful, you’ll see an AI response from Port ready to be passed to subsequent nodes.

---

⚙️ API Endpoints Implemented

Note

As of November 2025, only AI prompt endpoints are supported. Contributions for additional API paths are welcome.

Method	Path	Purpose	Path Params	Body (required)	Optional Params	Notes	Success Response
POST	/v1/agent/:agentIdentifier/invoke	Invoke a specific agent	agentIdentifier	—	context, prompt, labels, provider, model	Providers: openai | azure-openai | anthropic | bedrock | port Models: gpt-5 | claude-sonnet-4-20250514	JSON agent result
POST	/v1/ai/invoke	General-purpose AI interaction	—	invocation_identifier, userPrompt, tools	labels, provider, model, systemPrompt, executionMode	Execution mode: Automatic | Approval Required	JSON result
GET	/v1/ai/invoke/:invocation_identifier	Fetch invocation result	invocation_identifier	—	—	—	{ ok: boolean, result: object }

---

🤝 Contributing

Contributions are welcome!

1. Fork the repo
2. Create a new branch (feat/awesome-thing)
3. Make your changes
4. Submit a PR with a clear description of your update

Please include tests and examples when possible.

---

🪪 License

MIT © Port Labs

---

📦 Example Usage (code)

// Example: Calling a Port AI node programmatically via n8n API

const axios = require('axios');

const res = await axios.post('https://n8n.local/api/v1/workflows/1/execute', {
  userPrompt: 'Summarize this README for me!',
  provider: 'openai',
  model: 'gpt-5'
});

console.log(res.data);

---

🏷️ Metadata

Field	Value
Package name	@port-labs/n8n-nodes-portio-experimental
Version	1.0.0-alpha.5 (experimental release)
Node type	Custom n8n nodes
Focus	Port AI integrations
Status	Experimental / early access