Workflow

Overview

Blocks Workflow is a visual automation builder for connecting apps and orchestrating tasks. Design workflows with triggers, actions, and conditional paths using a drag-and-drop interface.

Dynamic, Flexible, Little to No Code

info

Developer Preview Release

Core functionality is available. Advanced features like conditionals, loops, and built-in functions are coming soon.

---

Core Concepts

Nodes

Every step in a workflow is called a node. Each node is specialized for a specific type of task. You can add nodes, configure them, and connect them in a logical sequence to accomplish your desired outcomes.

Trigger Nodes

Triggers are the initiation points of a workflow. A workflow executes only when the trigger's conditions are met. Every workflow must start with a trigger node.

Available Triggers:

• Webhook
• Blocks Email Trigger

Action Nodes

Action Nodes perform tasks such as calling a service, fetching data, sending emails, or making HTTP requests.

Available Actions:

• AI Agent
• Send Email
• HTTP Request

Executions

Each run of a workflow is considered an execution:

• Node Execution: Running an individual node (typically during configuration and testing)
• Workflow Execution: Running the entire workflow from trigger to completion

Each workflow execution has its own unique history that can be viewed from the Executions tab.

Execution Path

Workflows are executed starting from the trigger node and following connected paths:

• Each connected node is executed in the order of the connections
• Each node can access the data output from all nodes that executed before it in the chain
• Nodes execute sequentially following the connection flow

Expressions

Expressions are used to reference data and variables within your workflow. They follow a specific syntax and are always wrapped in double curly braces: {{}}

Expressions allow you to:

• Reference output from previous nodes
• Pass dynamic data between nodes
• Configure node parameters with values from earlier steps

See the Writing Expressions section for detailed syntax.

---

Available Nodes

Webhook

Generates a unique HTTP webhook URL that can be called to execute the workflow.

Configuration:

• Automatically generates a unique URL upon workflow creation/activation
• Copy the webhook URL and use it in external systems

Current Behavior:

• The webhook must be called as an HTTP POST request
• The request body must contain a RAW JSON payload
• The webhook returns an instant HTTP 200 response with a message indicating the workflow has been queued
• The workflow executes asynchronously after the response is sent

Output:
The webhook node outputs the entire JSON payload received in the POST request, which can be referenced by subsequent nodes.

Current Constraints:

• Only POST requests are supported
• Only RAW JSON request bodies are supported
• No support for GET requests, headers, query parameters, or other body formats

Planned Features:

• GET request support
• Custom headers in requests
• Query parameters
• Multiple body formats (form-data, x-www-form-urlencoded)
• "Respond to webhook" node to send custom responses
• "Respond after last node" option to return workflow results

---

Blocks Email Trigger

Triggers the workflow when an email is received in a connected inbound mailbox through Blocks Email Service.

How It Works:

• Connects to your mailbox via IMAP
• Polls the mailbox at regular intervals (configurable)
• Triggers the workflow when a new email is detected

Output Parameters:

Parameter	Description
ItemId	Unique identifier for the email item
MessageId	Email message ID from headers
MailServerConfigurationId	ID of the connected mailbox configuration
Subject	Email subject line
From	Sender email address
To	Recipient email address(es)
Body	Email body content
Status	Processing status of the email
Error	Error message if any occurred during retrieval
Date	Email date/timestamp
RawMime	Raw MIME content of the email
IsInBound	Boolean indicating if email is inbound (always true for this trigger)

Current Constraints:

• Only mailboxes with IMAP support can be connected
• Uses polling mechanism (not real-time push)

Example Usage:
Trigger workflows based on incoming support emails, form submissions, or automated notifications.

---

AI Agent

Allows you to integrate Blocks AI Agents into your workflow. AI Agents are custom intelligent agents you build within the Blocks platform.

How It Works:

• Select a pre-configured AI Agent from your Blocks workspace
• Provide input to the agent (can be static text or dynamic data from previous nodes)
• The agent processes the input according to its configuration, which may include:
  • Knowledge retrieval from connected data sources
  • Tool calling and function execution
  • Natural language processing
  • Data analysis and transformation
• The agent's response becomes available as output for subsequent nodes

Use Cases:

• Analyze email content and extract key information
• Generate responses or summaries
• Make intelligent decisions based on workflow data
• Process natural language inputs
• Query knowledge bases or documents

Example:
Use an AI Agent to analyze incoming support emails, categorize them, and extract customer intent before routing to the appropriate action.

---

Send Email

Sends emails using the Blocks Email Service with pre-configured email templates.

How It Works:

• Email templates are created and managed separately in Blocks Template Management service
• Templates support dynamic variables (placeholders)
• In the workflow, you select which template to use
• You provide values for the template variables dynamically based on workflow logic

Configuration:

• Template: Select from your available email templates
• Recipient: Specify email address(es) - can be dynamic using expressions
• Template Variables: Map workflow data to template variables

Example:
If your template has variables {{customer_name}} and {{order_id}}, you can populate these with data from previous workflow nodes like webhook input or AI Agent output.

Current Constraints:

• Templates must be created in Blocks Template Management service before use
• Template variable structure is defined in the template, not in the workflow

---

HTTP Request

Make HTTP requests to external APIs and services from your workflow.

Supported Methods:

• GET
• POST
• PUT
• PATCH
• DELETE

Supported Parameters:

Parameter Type	Description
Query Params	URL query parameters (e.g., ?key=value&foo=bar)
Headers	Custom HTTP headers for authentication, content-type, etc.
Body	Request body in JSON format only (for POST, PUT, PATCH)
Path Params	Dynamic URL path segments using expressions directly in the URL field

Example URL with Expressions:

https://api.example.com/users/{{$json.output.userId}}/orders

Use Cases:

• Call external APIs to fetch or send data
• Integrate with third-party services
• Post data to databases or analytics platforms
• Trigger actions in other systems

Current Constraints:

• Only JSON body format is supported
• No built-in OAuth or advanced authentication (use headers for API keys/tokens)

Planned Features:

• OAuth authentication support
• Multiple body formats (form-data, XML, etc.)
• Certificate-based authentication

---

Writing Expressions

Expressions allow you to dynamically reference data from previous nodes in your workflow. They are wrapped in double curly braces {{}}.

Syntax Rules

Referencing the Immediate Previous Node

When you want to reference output from the node directly connected before the current node:

{{$json.output.<param_name>}}

Breakdown:

• $json - Indicates we're accessing JSON data from the previous node
• .output - References the output section of that node
• .<param_name> - The specific parameter/key you want to access

Examples:

{{$json.output.email}}  
{{$json.output.Subject}}  
{{$json.output.userId}}

Referencing Any Previous Node (by Name)

When you want to reference output from a specific node that is not the immediate previous node:

{{$node["<​Node Name>"].json.output.<param_name>}}

Breakdown:

• $node["<​Node Name>"] - Targets a specific node by its name
• .json - References the JSON data from that node
• .output - References the output section
• .<param_name> - The specific parameter/key you want to access

Examples:

{{$node["Webhook"].json.output.customerEmail}}  
{{$node["Email Trigger"].json.output.Subject}}  
{{$node["AI Agent"].json.output.category}}

Expression Examples

Example 1: Using webhook data in an email

Recipient: {{$node["Webhook"].json.output.email}}  
Template Variable (customer_name): {{$node["Webhook"].json.output.name}}

Example 2: Using AI Agent output in HTTP request

URL: https://api.crm.com/contacts/{{$json.output.contactId}}  
Body:   
{  
  "status": "{{$json.output.category}}",  
  "notes": "{{$json.output.summary}}"  
}

Example 3: Passing email trigger data to AI Agent

AI Agent Input: Analyze this email and categorize it: {{$json.output.Body}}

Important Notes

• Node names must match exactly (case-sensitive)
• Parameter names must match the output structure of the node
• Use quotes around node names when using $node["Name"] syntax
• No built-in functions are currently available (string manipulation, date formatting, etc.) - coming in future releases

---

Workflow Management

Activation & Deactivation

• Workflows must be activated to run automatically when triggered
• Deactivated workflows will not execute, even if trigger conditions are met
• You can toggle activation status from the workflow settings
• Manual testing is available regardless of activation status

Execution History

• Every workflow execution is logged with a complete history
• View execution details in the Executions tab
• See the data passed through each node
• Debug failures by examining node outputs
• No current limits on execution history retention

Current Limitations

• No execution count limits per day/month
• No limits on number of nodes per workflow
• No automatic execution timeout (workflows run until completion)
• No workflow versioning (changes are applied directly)
• No rollback capability

Planned Features:

• Workflow versioning
• Execution limits based on plan tiers
• Configurable timeout settings
• Rollback to previous versions

---

Upcoming Features

The following features are planned for future releases:

Conditional Logic

• IF/THEN nodes for branching logic
• Switch nodes for multi-path routing
• Execute different paths based on conditions

Loops & Iteration

• Loop over arrays of data
• Process multiple items sequentially or in parallel
• Batch operations

Error Handling

• Try-catch mechanisms
• Fallback paths for failed nodes
• Retry logic with configurable attempts

Data Transformation

• Dedicated nodes for data manipulation
• Filtering, mapping, and reducing arrays
• Date/time formatting
• String operations

Built-in Expression Functions

• String manipulation (concatenate, split, replace, etc.)
• Mathematical operations
• Date/time functions
• Array operations (filter, map, find, etc.)
• JSON parsing and manipulation

Additional Triggers

• Schedule/Cron triggers for time-based automation
• Database triggers for data changes
• Form submission triggers

Enhanced HTTP Request

• OAuth 2.0 authentication
• Multiple body formats (form-data, XML, GraphQL)
• File upload support
• Certificate-based authentication

Webhook Enhancements

• GET request support
• Custom response handling
• Synchronous execution with result response
• Header and query parameter support

---

Support & Feedback

This is a Developer Preview release. Your feedback is invaluable for improving Blocks Workflow.

For questions, issues, or feature requests, please contact the Blocks platform support team.

---

Last Updated: Developer Preview Release