Human task node - MagOneAI

Purpose

The Human Task node pauses the workflow and waits for a human to provide input or approval before continuing. Essential for human-in-the-loop patterns, Human Task nodes ensure that critical decisions have human oversight. Human Task nodes bridge the gap between AI automation and human judgment. They enable workflows that leverage AI speed and consistency while retaining human oversight for high-stakes decisions.

How it works

When execution reaches a Human Task node, the workflow pauses durably while waiting for human input.

Workflow reaches Human Task

Execution arrives at the Human Task node. The workflow prepares to pause.

Execution pauses

The workflow execution pauses. Temporal maintains the complete state durably, so no data is lost.

Task appears in MagOneAI Hub

The task appears in MagOneAI Hub’s task queue for the assigned user or role. The user sees:

Task description
Context data from previous nodes
Available response options
Attachments or related documents

Human reviews and responds

The assigned user reviews the context, makes a decision, and submits their response through MagOneAI Hub.

Workflow resumes

The human’s response is captured and stored in the variable store. Workflow execution resumes from exactly where it paused.

Workflow continues

Subsequent nodes can access the human’s response and continue processing based on the decision.

Human Task nodes leverage Temporal’s durable execution. The workflow can wait for hours, days, or even weeks for human input without losing state or consuming compute resources. When the human responds, execution resumes instantly.

Configuration

Configure a Human Task node to present the right information and collect the right response.

Task description

Write a clear description of what the human needs to review or decide. The description appears as the task title in MagOneAI Hub. Good task descriptions:

“Review extracted invoice data for accuracy”
“Approve high-value purchase request: AED ”
“Validate compliance findings before finalizing report”
“Authorize payment of to ”

Poor task descriptions:

“Review” (too vague)
“Check this” (no context)
“Agent output validation” (too technical)

Assignee

Specify who should receive the task:

Specific user
Role
Dynamic assignment

Assign to a specific user by email or user ID.

{
  "assignee_type": "user",
  "assignee": "manager@company.com"
}

Use when: You know exactly who should handle this task.

Assign to anyone with a specific role.

{
  "assignee_type": "role",
  "assignee": "compliance_officer"
}

Use when: Any member of a team can handle the task.

Assign based on data from previous nodes.

{
  "assignee_type": "user",
  "assignee": "{{trigger.input.assigned_manager}}"
}

Use when: The assignee depends on workflow context (e.g., account manager for a specific customer).

Context data

Provide the human with context from previous nodes. This is the information they need to make an informed decision. Example: Invoice approval context

{
  "invoice_number": "{{document_agent.output.invoice_number}}",
  "vendor": "{{document_agent.output.vendor_name}}",
  "amount": "{{document_agent.output.total_amount}}",
  "currency": "{{document_agent.output.currency}}",
  "extracted_data": "{{document_agent.output}}",
  "risk_assessment": "{{risk_agent.output}}",
  "compliance_check": "{{compliance_agent.output}}",
  "document_url": "{{trigger.input.document_url}}"
}

The user sees this data in MagOneAI Hub formatted for easy review.

Approval options

Define the response options available to the human.

Simple approve/reject
Multiple options
Custom input

Binary decision with optional comments.

{
  "options": [
    {"value": "approved", "label": "Approve"},
    {"value": "rejected", "label": "Reject"}
  ],
  "allow_comments": true,
  "require_comments_on_reject": true
}

More than two choices.

{
  "options": [
    {"value": "approve", "label": "Approve"},
    {"value": "reject", "label": "Reject"},
    {"value": "request_more_info", "label": "Request More Information"},
    {"value": "escalate", "label": "Escalate to Director"}
  ],
  "allow_comments": true
}

Allow the human to provide structured data.

{
  "response_type": "form",
  "fields": [
    {
      "name": "decision",
      "type": "select",
      "label": "Decision",
      "options": ["Approve", "Reject", "Request Changes"]
    },
    {
      "name": "amount_adjustment",
      "type": "number",
      "label": "Adjusted Amount (if different)",
      "optional": true
    },
    {
      "name": "notes",
      "type": "text",
      "label": "Notes",
      "required": true
    }
  ]
}

Timeout

Configure what happens if the human doesn’t respond within a specified time. Timeout settings:

Duration — How long to wait (e.g., 24 hours, 3 days, 1 week)
Action on timeout — What happens when time expires
- Take default action (approve or reject)
- Escalate to another user
- Fail the workflow
- Execute a fallback branch

Example timeout configuration:

{
  "timeout_duration": "24h",
  "timeout_action": "escalate",
  "escalate_to": "director@company.com",
  "escalation_message": "This task has been waiting for 24 hours and requires immediate attention."
}

Escalation

Define escalation paths for time-sensitive tasks. Escalation chain:

Initial assignee → Wait 24 hours → Escalate to manager
Manager → Wait 12 hours → Escalate to director
Director → Wait 6 hours → Auto-approve with notification

Escalation configuration:

{
  "escalation_chain": [
    {
      "assignee": "{{trigger.input.assigned_user}}",
      "timeout": "24h"
    },
    {
      "assignee": "{{trigger.input.manager}}",
      "timeout": "12h"
    },
    {
      "assignee": "director@company.com",
      "timeout": "6h"
    }
  ],
  "final_action": "auto_approve",
  "final_action_notification": "director@company.com"
}

Use cases

Human Task nodes enable sophisticated human-in-the-loop workflows.

Financial approvals

Scenario: Approve invoices above a threshold before payment.

Workflow:
  1. Extract invoice data (Agent)
  2. Validate against PO (Agent)
  3. Check budget availability (Tool)
  4. Condition: amount > 10000
     ├─ True: Human Task: "Finance manager approval"
     └─ False: Auto-approve
  5. Process payment (Tool)
  6. Send confirmation (Tool)

Human Task configuration:

{
  "description": "Approve invoice payment: {{vendor}} - {{amount}} {{currency}}",
  "assignee": "finance_manager@company.com",
  "context": {
    "invoice_number": "{{invoice_number}}",
    "vendor": "{{vendor_name}}",
    "amount": "{{amount}}",
    "currency": "{{currency}}",
    "po_number": "{{po_number}}",
    "budget_remaining": "{{budget_check.output.remaining}}",
    "invoice_url": "{{document_url}}"
  },
  "options": [
    {"value": "approved", "label": "Approve Payment"},
    {"value": "rejected", "label": "Reject Invoice"}
  ],
  "timeout": "48h",
  "timeout_action": "escalate",
  "escalate_to": "cfo@company.com"
}

Content review before publication

Scenario: Review AI-generated content before publishing to customers.

Workflow:
  1. Generate content (Agent with RAG)
  2. Fact-check against knowledge base (Agent)
  3. Check brand guidelines (Agent)
  4. Human Task: "Content review and approval"
  5. Condition: approved
     ├─ True: Publish content (Tool)
     └─ False: Return to drafts
  6. Send publication notification (Tool)

Human Task configuration:

{
  "description": "Review article before publication: {{article_title}}",
  "assignee_role": "content_reviewer",
  "context": {
    "title": "{{article_title}}",
    "content": "{{generated_content}}",
    "target_audience": "{{audience}}",
    "publication_date": "{{scheduled_date}}",
    "fact_check_results": "{{fact_checker.output}}",
    "brand_check_results": "{{brand_checker.output}}"
  },
  "options": [
    {"value": "approve", "label": "Approve for Publication"},
    {"value": "request_changes", "label": "Request Changes"},
    {"value": "reject", "label": "Reject"}
  ],
  "allow_editing": true,
  "timeout": "4h"
}

Data validation

Scenario: Validate AI-extracted data before committing to database.

Workflow:
  1. Extract data from document (Agent with vision)
  2. Validate format and completeness (Agent)
  3. Condition: confidence < 0.9
     ├─ True: Human Task: "Validate extracted data"
     └─ False: Auto-approve
  4. Save to database (Tool)
  5. Send confirmation (Tool)

Human Task configuration:

{
  "description": "Validate extracted data: {{document_type}}",
  "assignee_role": "data_entry_specialist",
  "context": {
    "document_image": "{{document_url}}",
    "extracted_data": "{{extractor.output}}",
    "confidence_score": "{{extractor.output.confidence}}",
    "validation_issues": "{{validator.output.issues}}"
  },
  "response_type": "form",
  "fields": [
    {"name": "name", "type": "text", "label": "Full Name", "default": "{{extracted_data.name}}"},
    {"name": "id_number", "type": "text", "label": "ID Number", "default": "{{extracted_data.id}}"},
    {"name": "expiry_date", "type": "date", "label": "Expiry Date", "default": "{{extracted_data.expiry}}"},
    {"name": "verified", "type": "checkbox", "label": "I have verified this data is correct"}
  ],
  "timeout": "1h"
}

Compliance escalation

Scenario: Escalate high-risk compliance findings to legal team.

Workflow:
  1. Analyze document for compliance (Agent)
  2. Assess risk level (Agent)
  3. Condition: risk_level == "high"
     ├─ True: Human Task: "Legal review required"
     └─ False: Auto-process
  4. Generate compliance report (Agent)
  5. Save report (Tool)

Human Task configuration:

{
  "description": "URGENT: High-risk compliance issue requires legal review",
  "assignee_role": "legal_counsel",
  "priority": "high",
  "context": {
    "document_type": "{{document_type}}",
    "risk_level": "{{risk_agent.output.risk_level}}",
    "findings": "{{compliance_agent.output.findings}}",
    "recommendations": "{{compliance_agent.output.recommendations}}",
    "document_url": "{{document_url}}",
    "customer_id": "{{customer_id}}"
  },
  "options": [
    {"value": "approve", "label": "Approve with Conditions"},
    {"value": "reject", "label": "Reject - Non-compliant"},
    {"value": "request_more_info", "label": "Request Additional Information"}
  ],
  "require_comments": true,
  "timeout": "4h",
  "escalate_to": "chief_legal_officer@company.com"
}

Best practices

Provide rich context

Give humans all the information they need to make an informed decision. Include:

Summary of what happened before
Key data points
AI analysis results
Links to original documents
Relevant history

Set realistic timeouts

Consider business hours, time zones, and realistic response times. Don’t set a 1-hour timeout if the person might be asleep.Good timeout examples:

Urgent approval: 4 hours with escalation
Standard approval: 24 hours
Non-urgent review: 3 days

Use clear action labels

Instead of generic “Yes/No”, use specific labels: “Approve Payment”, “Reject Invoice”, “Request Changes”.

Make it easy to approve

For low-risk tasks, make approval the default and easiest action. For high-risk tasks, require explicit affirmation.

Require comments on rejection

When humans reject something, require them to explain why. This provides valuable feedback and audit trails.

Build escalation chains

Don’t let tasks sit indefinitely. Build escalation chains that ensure timely resolution.

Track approval metrics

Monitor approval times, approval rates, and who approves what. Use this data to optimize workflows and identify training needs.

Use Human Task nodes strategically. Not every decision needs human approval. Reserve human oversight for high-stakes decisions, low-confidence AI outputs, and situations where human judgment is genuinely needed.

Advanced patterns

Multi-stage approvals

Require multiple approvals in sequence:

Workflow:
  1. Process request (Agent)
  2. Human Task: "Manager approval"
  3. Condition: approved AND amount > 50000
     ├─ True: Human Task: "Director approval"
     └─ False: Continue
  4. Execute action (Tool)

Parallel approvals

Require multiple approvals simultaneously:

Workflow:
  1. Process request (Agent)
  2. Parallel Node:
     ├─ Branch 1: Human Task: "Finance approval"
     ├─ Branch 2: Human Task: "Legal approval"
     └─ Branch 3: Human Task: "Compliance approval"
  3. Condition: All approved
     ├─ True: Execute
     └─ False: Reject

Conditional approval

Only require approval under certain conditions:

Workflow:
  1. Process document (Agent)
  2. Assess risk (Agent)
  3. Condition: risk_score > 0.7 OR amount > threshold
     ├─ True: Human Task: "Approval required"
     └─ False: Auto-approve
  4. Continue processing

Approval with agent augmentation

Agent assists human decision-making:

Workflow:
Analyze request (Agent)
Research relevant precedents (Agent with RAG)
Generate recommendation (Agent)
Human Task: "Final decision" (with AI recommendation)
Process based on human decision

Integration with MagOneAI Hub

Human Task nodes integrate seamlessly with MagOneAI Hub’s task management interface.

Task queue

Tasks appear in the assignee’s queue with:

Priority indicators — Urgent, high, normal, low
Age indicators — How long the task has been waiting
Context preview — Quick view of key information
Filters — Filter by priority, age, workflow type

Task detail view

When a user opens a task, they see:

Full context — All data provided by the workflow
Attachments — Documents, images, links
Timeline — What happened before this task
Related tasks — Other tasks in the same workflow
Response options — Clear action buttons

Notifications

Users receive notifications when:

A new task is assigned to them
A task is approaching its timeout
A task has been escalated to them
Someone comments on their task

Mobile access

Human Task nodes work on mobile devices through MagOneAI Hub mobile app:

Push notifications for new tasks
Full context display optimized for mobile
Quick approve/reject actions
Camera integration for document review

Responding programmatically (API & webhooks)

Human tasks don’t have to be answered from the Hub UI. A workflow started through the webhook/API trigger can pause on a Human Task and be resumed by an external system — so you can build approval steps into integrations where the approver lives in another application. The flow mirrors the in-app experience over signed (HMAC-authenticated) API calls:

Trigger the workflow

Call the workflow’s webhook endpoint. The response includes a poll token and poll URL you use to follow execution.

Detect the pending task

When the workflow reaches a Human Task, its status becomes waiting for input and polling surfaces the pending human-task IDs.

Read the task

Fetch the task to get its questions, options, and current status — the same content a Hub user would see.

Submit the response

Post the answers back to the task. The platform validates them against the task’s questions, marks the task complete, and signals the workflow to resume. Submitting twice is safe — a task that’s already answered isn’t re-signalled.

Get the result

By default the submit call returns immediately and you poll for completion (including any new human task the workflow pauses on next). Optionally you can ask the call to wait briefly for the workflow to settle and return the outcome inline.

A webhook caller may only read and answer tasks belonging to executions it started, within the same project and any allowed-use-case restrictions on its API key. Requests outside that scope return a not-found error so task existence is never leaked.

This works for top-level executions as well as tasks inside schedules, sub-workflows, and foreach branches. Ready-to-paste request snippets (curl, Python, Node) are available in the API-key credentials screen in Studio.

Next steps

Condition node

Route to human tasks based on conditions

Parallel node

Require multiple parallel approvals

Agent node

Use agents to prepare context for human review

Memory system

Access human task responses in subsequent nodes

​Purpose

​How it works

​Configuration

​Task description

​Assignee

​Context data

​Approval options

​Timeout

​Escalation

​Use cases

​Financial approvals

​Content review before publication

​Data validation

​Compliance escalation

​Best practices

​Advanced patterns

​Multi-stage approvals

​Parallel approvals

​Conditional approval

​Approval with agent augmentation

​Integration with MagOneAI Hub

​Task queue

​Task detail view

​Notifications

​Mobile access

​Responding programmatically (API & webhooks)

​Next steps

Condition node

Parallel node

Agent node

Memory system

Purpose

How it works

Configuration

Task description

Assignee

Context data

Approval options

Timeout

Escalation

Use cases

Financial approvals

Content review before publication

Data validation

Compliance escalation

Best practices

Advanced patterns

Multi-stage approvals

Parallel approvals

Conditional approval

Approval with agent augmentation

Integration with MagOneAI Hub

Task queue

Task detail view

Notifications

Mobile access

Responding programmatically (API & webhooks)

Next steps