When to Use the Task Kanban
Use the Kanban when you need to:- Create tasks in natural language
- Generate AI execution plans with estimates
- Approve and execute tasks
- Respond to agent questions requiring clarification
- Track execution progress across environments and worker queues
- Manage failed, running, or completed tasks
- Review all historical executions associated with your workspace
Creating a Task
Task creation is intentionally simple, users describe what they want to achieve, and Kubiya handles the rest. Here is the step-by-step flow.1. Open “Create Task”
You’ll see a configuration form with:- A large natural-language input box
- Priority selector
- Optional advanced settings
2. Describe the Task
Enter what you want done, exactly as you would explain it to a teammate. Good prompts include:- What outcome is expected
- Requirements
- Constraints
- Tools, systems, or environments involved
- “Create a responsive landing page for our authentication product with a hero section, features list, and sign-up form.”
- “Deploy the new API version to staging and run smoke tests.”
- “Analyze cloud spend for the last 30 days and flag high-impact cost anomalies.”
3. Pick a Priority Level
Select:- Low
- Medium
- High
- Critical
4. (Optional) Open Advanced Options
Advanced Options let you anchor the task into your execution environment:Link to Project
Assign the task to an existing Project so it appears in that project’s backlog and Kanban board.Execution Environment
Select the environment where steps should run (e.g., development, staging, production).Worker Queue
Choose which queue will execute the steps. If not selected, Kubiya chooses the most suitable queue.5. Generate the Plan
Click Generate Plan. Your task is submitted and moves to the Planning column. The AI interprets your description and determines:- What exactly needs to be done
- Which systems, repos, or resources are involved
- The optimal executor (team or agent)
- Required steps, sequence, and dependencies
6. Review & Approve the Generated Plan
The plan includes:- Story points
- Engineering hours estimate
- Confidence level
- Risks and assumptions
- Requirements
- Recommended Team/Agent
- Execution context (Environment + Worker Queue)
- Detailed step-by-step sequence
- Approve & Execute → Task begins execution
- Send Feedback & Refine → Plan is regenerated
- Cancel → Task is discarded
7. Execution Begins
After approval:- The task enters Pending / Running depending on queue capacity
- A worker picks up execution steps
- Logs stream in real time
- Policies are checked before each step
- Monitor logs
- Cancel the execution
- Track step-level progress
8. Completion, Failure, or Manual Closure
Execution results in one of the following:- Completed: all steps executed successfully
- Failed: errors, missing prerequisites, policy violations
- Manually Closed: user terminates or abandons the task
Task Progress Flow
This describes the typical flow tasks go through on the Kanban.1. Planning
Task has been submitted; AI is generating the plan.2. Requires Input
The agent needs clarification before continuing. Users supply missing details and planning resumes.3. Ready for Review
The execution plan is complete. User decides to:- Approve
- Refine
- Cancel
4. Executing
The plan is approved and running. Shows:- Step-by-step activity
- Real-time logs
- Assigned agent/team
- Environment + Queue
- Status of each step
5. Completed / Failed / Closed
Final state depending on outcome:- Completed – success
- Failed – error, no worker capacity, policy rejection
- Closed – terminated manually or abandoned
CLI Usage
Use the CLI to list, inspect, monitor, and control all task executions. Executions represent any task that has been approved and sent to an agent or team for processing.1. List Executions
Lists all executions with optional filters for status, agent, team, or type. Commands| Flag | Description |
|---|---|
--status | Filter by status: pending, running, completed, failed |
--agent | Filter by agent ID |
--team | Filter by team ID |
--type | Filter by execution type: agent or team |
--limit | Limit result count (default 50) |
2. Get Execution Details
Displays full execution metadata including plan, steps, environment, and final results. Commands3. Stream Execution Logs
Streams real-time logs and events for an active execution. Use Ctrl+C to exit log streaming. Commands4. Cancel Executions
Cancels a running execution. Kubiya attempts a graceful shutdown and halts remaining steps. CommandsBest Practices
- Describe tasks with outcomes, requirements, and constraints
- Approve plans only after reviewing steps and risk notes
- Respond promptly if agents request clarification
- Cancel tasks that hang or are no longer needed
- Use CLI streaming for debugging failing or long-running tasks
- Check policies if tasks fail unexpectedly
- Ensure active workers are connected for real-time execution