Automations are built by connecting nodes together. Each node performs a specific function—from
taking inputs to querying data to sending notifications.
The Control Node is the starting point of every automation. All steps flow from this anchor point,
except for input nodes. While it’s not a building block you can add or duplicate, it provides
essential controls for running and scheduling your automation. Every automation has exactly one
Control Node.Actions:
Action
Description
Run
Execute the automation immediately with the current config
When you connect other nodes to your automation (like Input, Slack, or Email nodes), the Control
Node’s configuration panel will show fields for those connected nodes. This is where you:
The Input Node is typically the starting point of your automation. It defines what data users must
provide when triggering the workflow.Use cases:
Accept a wallet address to analyze
Take a date range for reporting
Receive a list of tokens to track
Configuration: | Field | Description | |-------|-------------| | Name | Variable name to
reference in other nodes (e.g., wallet_address) | | Type | Data type: Text, Number, Date,
Boolean, or Array | | Description | Help text shown when running the automation | | Default
| Optional default value | | Required | Whether the input must be provided |Example:
The Conditional Node lets you create if/then logic in your automation. Route data down different
paths based on values from previous nodes.Use cases:
Only send alerts if a threshold is exceeded
Skip email if query returns no results
Route to different agents based on input type
Configuration: | Field | Description | |-------|-------------| | Condition | Expression that
evaluates to true/false | | Then | Path to follow if condition is true | | Else | Path to
follow if condition is false |Example:
The Transform Node manipulates data as it flows through your automation. Use it to format outputs,
extract fields, or combine results from multiple sources.Use cases:
Format numbers as currency
Extract specific columns from query results
Combine data from multiple queries
Convert data types
Configuration: | Field | Description | |-------|-------------| | Input | Data to transform |
| Operation | Type of transformation | | Output | Resulting data structure |Example:
Agent Nodes use AI to analyze, summarize, or transform data from previous nodes.Use cases:
Summarize query results in natural language
Generate insights from data patterns
Format data for human-readable reports
Answer questions about the data
Configuration: | Field | Description | |-------|-------------| | Agent | Which agent to use
| | Prompt | Instructions for what the agent should do | | Input | Data to pass to the agent
|Example prompt:
Analyze this wallet's transaction history and provide:1. A summary of activity patterns2. Notable large transactions3. Most frequently interacted protocolsKeep the response under 300 words.
The sidebar shows available agents. Select one based on your use case—different agents have
different skills and expertise.