​
Overview

Nodes are the building blocks of a workflow in Pathlit. Each node performs a specific task or operation, and they can be connected to create a sequence of actions. By combining different nodes, users can design custom workflows to automate tasks, process data, and generate insights.

Each node has its own set of configuration parameters, inputs, and outputs. The configuration parameters define how the node behaves, while the inputs and outputs determine how data flows between nodes in the workflow. Understanding the purpose and functionality of each node is essential for creating effective workflows in Pathlit.

​
Types of Nodes

Nodes in Pathlit are categorized based on their functionality and purpose. Here are some common types of nodes you may encounter:

​
Triggers

Trigger nodes initiate the execution of a workflow based on specific events or conditions. Examples include:

  • Scheduler node for time-based triggers
  • Gmail Receive for email-based triggers
  • Slack Trigger for Slack message-based events
  • Bitbucket Event Receive for code repository events

​
Core Flow

Core Flow nodes manage the fundamental flow and data manipulation within your workflow. These include:

  • Input node for accepting user input
  • Simple Router for conditional branching
  • Text Variable for managing text data

​
AI & Agents

AI & Agents nodes leverage artificial intelligence and specialized agents to perform complex tasks:

  • LLM Node for direct language model interactions
  • Agent Node for create autonomous AI system that uses tools to complete tasks
  • Agent Swarm for multi-agent collaboration
  • Text-to-Image generation
  • Summarizer for text summarization

​
Integrations

Integration nodes connect your workflow with external services and platforms. These are grouped by service:

​
Google Workspace

  • Gmail (Send, Receive, Agent)
  • Google Drive Input
  • Google Sheets Agent
  • Google Calendar Agent
  • Google Tasks Agent

​
Communication Platforms

  • Slack (Message Send, Trigger)
  • Outlook (Send, Receive)

​
Development Tools

  • Bitbucket (Agent, PR Input, Event Receive)
  • Jira (Issue Input, Output, Receive)

​
Productivity Tools

  • Notion Page Output
  • Airtable (Input, Output)
  • LinkedIn Output

​
Subflows

Subflow nodes enable the creation of modular and reusable workflow components:

  • Workflow Invocation
  • Workflow Batch Invocation

​
Advanced

Advanced nodes provide specialized functionality for complex use cases:

  • Code Execution for custom Python code

​
Node Documentation

View the documentation for each node type to learn more about their configuration, expected inputs and outputs, use case examples, and error handling.

Agent

General-purpose AI agent for various tasks

Agent Swarm

Orchestrate multiple AI agents to collaborate on complex tasks

Airtable Input

Import data from Airtable into your workflow

Airtable Output

Export data to Airtable from your workflow

Audio Transcribe

Transcribe audio files into text

Bitbucket Agent

Interact with Bitbucket using an AI agent

Bitbucket Event Receive

Trigger workflows based on Bitbucket events

Bitbucket PR Input

Import data from Bitbucket pull requests

Code

Execute Python code within your workflow

Document Data

Process and manage document data

Embed Search

Perform vector similarity search

Google Calendar Agent

Manage calendar events using an AI agent

Global Variable

Store and manage workflow-wide variables

Gmail Agent

Automate email tasks using an AI agent

Gmail Receive

Trigger workflows from incoming emails

Gmail Send

Send emails from your workflow

Google Drive Input

Import files from Google Drive

Google Search

Perform Google searches within your workflow

Google Task Agent

Manage Google Tasks using an AI agent

Google Sheets Agent

Interact with Google Sheets using an AI agent

Input

Accept input data into your workflow

Jira Issue Input

Import Jira issues into your workflow

Jira Issue Output

Create or update Jira issues

Jira Issue Receive

Trigger workflows from Jira events

JSON

Process and transform JSON data

LinkedIn Output

Post content to LinkedIn

LinkedIn Scraper Agent

Utilize an agent to scrape LinkedIn posts, profiles, and companies

LLM

Interact with language models directly

Notion Page Output

Create or update Notion pages

OneDrive Input Node

Fetch and process files from Microsoft OneDrive

Outlook Receive

Trigger workflows from Outlook emails

Outlook Send

Send emails through Outlook

Output

Output data from your workflow

Scheduler

Schedule workflow executions

Simple Router

Control workflow execution paths

Slack Message Send

Send messages to Slack

Slack Trigger

Trigger workflows from Slack events

Sticky Note

Add notes and documentation to your workflow

Structured Output

Generate structured data from text

Summarizer

Create summaries of text content

Text Variable

Manage text data in your workflow

Text to Image

Generate images from text descriptions

Video Input

Process video files in your workflow

Web Extract

Extract data from web pages

Web Search

Search the web within your workflow

Workflow Batch Invocation

Run multiple workflow instances in parallel

Workflow Invocation

Call other workflows from your workflow