task_create
Create a new task to track work with subject, detailed markdown description, tags, metadata, dependencies, and parent-child relationships.
Instructions
Create a new task for tracking work
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| tags | No | Task tags | |
| owner | No | Who owns this task | |
| sources | No | URLs, references, documentation used for this task (e.g. ['https://docs.example.com', 'Claude Code Plan Mode analysis']) | |
| subject | Yes | Brief imperative title (e.g. 'Fix auth bug') | |
| metadata | No | Arbitrary key-value pairs (e.g. { pr: '#42', priority: 'high' }) | |
| blockedBy | No | Task IDs that must complete first | |
| parent_id | No | Parent ID, task parent ID whom task is related to | |
| project_id | No | Project key this task belongs to | |
| description | No | Detailed markdown description. When planning, document yourself regarding the task (e.g. search information online (use web_search tool if allowed), consult the codebase or ask further questions before delivering a plan). Preparing a plan can take several task update cycles until all the information is gathered. Structured description, details points when applicable: - Context: explain the why, and overall view how to solve the problem - Task details - Task location: where the task takes place (filepath or physical location, etc.) - Architecture: final structure of the expected output - Dependencies: explain what are required for delivery - Examples of the deliverable and expectations - Reasoning for the offered solution - Foreseen barriers and solution (from task context) - Implementation steps: real, commitable implementation steps, questions to ask the user for clarification and decision - Testing plan - Deployment plan - Sources and reference: explicitely list the sources used to design the plan, or say no sources have been used. |