Tasks & Assignments Guide

Overview

Tasks are the atomic units of work in Workunit. Each workunit breaks down into multiple tasks (typically 5-10, ranging from 2-20 depending on complexity) that can be tracked, assigned, and completed independently. Workunit's task system enables clear ownership and seamless collaboration across your entire team.

This guide covers everything you need to know about creating effective tasks, managing assignments, tracking progress, and organizing work for maximum team productivity.

Creating Tasks

Tasks can be created through the web interface or by AI assistants via MCP. Both methods support the full range of task properties and configurations.

Task Creation Basics

Creating a task requires only a title and parent workunit. All other properties are optional but recommended for effective tracking:

Via Web Interface:

Open the workunit where you want to add tasks
Click "Add Task" or "New Task" button
Fill in task properties in the form
Click "Create Task" to save

Via AI Assistant:

"Add a task to workunit 'User Authentication': Implement JWT token generation with 24-hour expiry"

AI models can create tasks directly via MCP, setting properties like priority, assignee, and estimates automatically.

Bulk Task Creation:

"Break down workunit 'Mobile App Redesign' into 8-10 specific implementation tasks with priorities"

AI assistants excel at generating comprehensive task lists from high-level goals, saving significant planning time.

Writing Effective Tasks

Good tasks are clear, actionable, and appropriately scoped. Follow these principles:

Start with Action Verbs

Begin task titles with verbs like Implement, Design, Write, Test, Review, Deploy, Fix, Analyze

Good

Implement user registration endpoint

Avoid

User registration

Be Specific and Measurable

Include concrete deliverables and success criteria in the task description

Good

Add email validation with regex pattern and error messages

Avoid

Improve validation

Keep Scope Manageable

Tasks should take 2-8 hours. Larger work should be broken into multiple tasks

Good

Write unit tests for auth middleware (10 test cases)

Avoid

Build complete authentication system

Add Context in Description

Include relevant technical details, constraints, and references in the description field

Example: "Use bcrypt with cost factor 12 for password hashing. Reference OWASP guidelines for secure storage. Ensure consistent timing to prevent timing attacks."

Task Properties

Tasks support a rich set of properties for tracking, organizing, and prioritizing work. Understanding these properties helps you manage tasks effectively.

Core Properties

Title

Brief, action-oriented description of the task (required)

Max 200 characters. Use action verbs and be specific.

Description

Detailed context, requirements, technical notes, and acceptance criteria (optional)

Markdown supported. Include links, code snippets, and implementation details.

Status

Current state of the task (default: todo)

todo in progress done blocked

Priority

Relative importance for sequencing work (default: normal)

critical high normal low

Assignee

Person responsible for completing the task (optional)

Tasks can be assigned to team members or left unassigned for the task pool.

Advanced Properties

Due Date

Target completion date for planning and deadline tracking

Soft deadline for planning. Use blocking status for hard dependencies.

Estimated Hours

Expected time to complete the task for capacity planning

Use for sprint planning and workload balancing.

Actual Hours

Time spent working on the task for tracking and retrospectives

Updated manually or via time tracking. Useful for improving future estimates.

Task Assignments

Assign tasks to team members for clear ownership and accountability, or leave them unassigned in a shared task pool. Assignments provide clarity about responsibility while allowing flexibility for dynamic workflows.

Assigning Tasks to Team Members

Assign tasks to specific team members for accountability and workload visibility:

Via Web Interface:

Open the task detail view
Click the "Assigned To" field dropdown
Select a team member from your organization
Task appears in their personal task list

Via AI Assistant:

"Assign task 'Write integration tests' to Sarah in workunit 'API Security'"

AI assistants can assign tasks via MCP when creating or updating tasks, streamlining workflow management.

Reassignment:

Tasks can be reassigned at any time. Previous assignee and new assignee both see the change. Use reassignment when priorities shift or expertise requirements change.

Unassigned Task Pool

Not all tasks need immediate assignment. The unassigned task pool provides flexibility for dynamic workflows:

When to Leave Tasks Unassigned:

• Tasks that depend on completing earlier work first
• Work that any qualified team member can pick up
• Tasks awaiting clarification or additional information
• Backlog items for future sprints or iterations
• Exploratory or research tasks without clear ownership yet

Benefits of Task Pool:

• Team members can self-assign based on availability and expertise
• AI models can pick up unassigned tasks that match their capabilities
• Reduces bottlenecks from over-assignment to specific people
• Enables more flexible, pull-based workflows

Task Status Management

Task status provides real-time visibility into work progress. Status updates sync instantly across all AI models and team members, enabling true parallel collaboration.

Task Status Lifecycle

Todo (Default)

Task is defined and ready to be started. Sits in backlog or task pool until someone begins work.

Transition: When work begins → in_progress

In Progress (Active)

Someone is actively working on this task. Signals to others that work is underway to prevent duplication.

Transition: When work finishes → done, or if blocked → blocked

Done (Completed)

Task is complete and meets acceptance criteria. Work can proceed to dependent tasks.

Final state: Usually permanent, but can reopen if issues discovered

Blocked (Impediment)

Task cannot proceed due to external dependency, missing information, or technical blocker. Add blocker details in comments.

Transition: When blocker resolved → in_progress or todo

AI-Driven Status Updates

AI assistants can update task status via MCP, enabling automatic progress tracking as they work:

End-to-end workunit execution:

"Implement workunit 'Email Verification Feature' and update task status as you complete each one"

AI works through all tasks sequentially, marking each in_progress → done as it completes them, providing real-time visibility into progress

Starting individual tasks:

"Mark task 'Implement JWT generation' as in progress"

AI updates status when beginning work, preventing duplicate effort by team members or other AI sessions

Completing tasks:

"Mark task 'Write unit tests' as done and add completion notes to AI context"

AI marks tasks complete and documents what was accomplished for future reference

Encountering blockers:

"Mark task 'Deploy to staging' as blocked - AWS credentials not configured"

AI documents blockers so humans can resolve issues and work can continue

Bulk status updates:

"Mark all completed tasks in workunit 'API Refactoring' as done"

AI can update multiple tasks efficiently based on actual completion state

Task Dependencies

Some tasks must be completed before others can begin. Workunit handles dependencies through task ordering and status management rather than explicit dependency links, keeping things simple while enabling complex workflows.

Managing Dependencies

Document in Task Description

Reference prerequisite tasks in the task description for clarity.

Example: "Add authentication middleware (requires: JWT generation task completed first)"

Use Blocked Status

Mark tasks as blocked when waiting for dependencies, making the constraint visible.

Example: Status: blocked, Comment: "Waiting for API schema design to be finalized"

Use Priority Levels

Set higher priority on prerequisite tasks to signal they should be completed first.

Example: "Design database schema" (high priority) before "Implement schema migration" (normal priority)

Communicate with Team

Use comments or workunit context to explain task dependencies. AI assistants and team members will follow documented sequences.

Task Organization

Effective task organization makes work easier to find, prioritize, and manage. Use tags and ordering to create structure without rigid processes.

Using Tags for Task Organization

Tags provide flexible categorization for filtering and grouping tasks across workunits:

By Technology Stack

frontend backend database devops

Helps team members find work matching their expertise

By Task Type

feature bug docs testing refactor

Clarifies the nature of work for planning and estimation

By Work Phase

planning implementation review deployment

Tracks where tasks are in your development workflow

By Priority Context

security performance accessibility mvp

Highlights important characteristics or requirements

Task Naming Conventions

Clear, consistent task naming helps team members quickly understand task scope and find related work:

Use Action Verbs

Start with verbs like Implement, Design, Write, Test, Review, Deploy, Fix, Refactor

Include Component Name

Reference the specific component or feature: "Implement user registration API endpoint"

Be Specific

Avoid vague titles. "Add email validation with error messages" beats "Improve validation"

Phase Prefixes (Optional)

Use prefixes for workflow clarity: "[Backend] Add JWT middleware" or "[Testing] Write integration tests"

Number Series (Optional)

For sequential work: "1. Design schema", "2. Implement migration", "3. Add test data"

Time Tracking & Estimates

Track estimated and actual time to improve planning accuracy and understand team capacity. Workunit keeps time tracking simple and non-invasive.

Estimated Hours (Planning)

Set during task creation or planning phase. Used for sprint capacity and workload balancing.

"Create task: Implement password reset flow (estimate: 4 hours)"

Actual Hours (Tracking)

Updated when task is completed. Compare with estimates to improve future planning.

"Mark task 'Implement password reset' as done (actual time: 5.5 hours)"

AI Assistance with Estimates

Ask AI models to suggest time estimates based on task complexity and historical data from similar tasks.

"Suggest time estimates for all tasks in workunit 'API Redesign' based on complexity"

Lightweight Philosophy

Time tracking is optional. Use it when helpful for planning, skip it when it creates unnecessary overhead. Workunit focuses on outcomes, not micromanagement.

Best Practices

Apply these principles to maximize task effectiveness and team productivity:

Keep Tasks Atomic

Each task should represent a single, complete unit of work. If a task has multiple distinct parts, split it into multiple tasks.

Write for Both Humans and AI

Task descriptions should be clear enough for AI assistants to understand context and detailed enough for humans to pick up where AI left off.

Update Status Frequently

Keep status current so team members and AI models have accurate visibility. Stale status creates confusion and duplicate work.

Add Context as You Learn

Update task descriptions with implementation details, gotchas, or constraints discovered during work. Future context helps everyone.

Use Priority Judiciously

Not everything is high priority. Reserve critical/high for truly urgent work. Normal priority should be the default.

Close Completed Tasks Promptly

Mark tasks done as soon as they meet acceptance criteria. Clean done/todo separation maintains clear focus.

Document Blockers Immediately

When stuck, mark blocked and document the issue. Transparency enables help and prevents wasted effort.

Review Tasks in Retrospectives

Compare estimates vs actuals, review what went well, identify patterns. Use learnings to improve future task planning.

Next Steps

Master task management and explore related Workunit features:

AI Features Guide

Learn how AI assistants can create, update, and manage tasks automatically

Core Concepts

Understand how tasks fit into workunits, assets, and context preservation

Team Collaboration

Set up your team for effective task assignment and collaborative workflows

Quick Start Guide

Create your first workunit with tasks in 10 minutes

Questions About Task Management?

Join our community discussions or check the documentation for detailed guidance on task workflows.

Documentation Community Discussions