If you're using Cursor IDE, add the rules to your User Rules to enable AI assistants to properly create and manage tasks:
Open Cursor Settings → User Rules
Copy the content from rules/.cursorrules (see below)
Paste it into your User Rules
Copy Cursor Rules
# Markdown Kanban & Roadmap - Cursor Rules
> **Note for AI Assistants**: These rules are suggested guidelines for AI assistants to understand how to properly create and manage tasks in the Markdown Kanban format. When working with `TASKS.md` files, follow these rules to ensure compatibility with the VS Code Markdown Kanban extension.
This file defines the rules for creating and managing tasks in the Markdown Kanban format that works with the VS Code Markdown Kanban extension.
## TASKS.md File Structure
The `TASKS.md` file must follow the Kanban format with these required sections:
- **Backlog**: Tasks that haven't been started yet
- **Doing**: Tasks currently in progress
- **Review**: Tasks awaiting review
- **Done**: Completed tasks
- **Paused**: Temporarily paused tasks
- **Notes**: General project notes (optional)
## Task Format
### Basic Structure
Each task must follow this format:
```markdown
### Task Name
- id: T-001
- tags: [tag1, tag2, tag3]
- priority: high|medium|low
- workload: Easy|Normal|Hard|Extreme
- milestone: sprint-26-1_1
- start: YYYY-MM-DD
- due: YYYY-MM-DD
- updated: YYYY-MM-DD
- completed: YYYY-MM-DD
- detail: ./tasks/T-001.md
- defaultExpanded: true|false
```
### Task Properties
#### Required
- **`id`**: Unique task identifier (format: `T-XXX` where XXX is a sequential number)
- Example: `T-001`, `T-002`, `T-010`
#### Optional
- **`tags`**: Array of tags for categorization
- Format: `[tag1, tag2, tag3]`
- Example: `[backend, frontend, ui, components]`
- Use descriptive and consistent tags
- **`priority`**: Priority level
- Values: `high`, `medium`, `low`
- **`workload`**: Estimated effort
- Values: `Easy`, `Normal`, `Hard`, `Extreme`
- **`milestone`**: Milestone or marker the task belongs to
- Suggested format: `sprint-year-month_number` (e.g., `sprint-26-1_1` for January 2026, sprint 1)
- Can use custom strings (e.g., `reconciliation-nn`, `timezone-utc`)
- **`start`**: Start date
- Format: `YYYY-MM-DD`
- Example: `2025-12-01`
- **`due`**: Due date
- Format: `YYYY-MM-DD`
- Example: `2026-01-15`
- **`updated`**: Last update date
- Format: `YYYY-MM-DD`
- Update whenever the task is modified
- **`completed`**: Completion date
- Format: `YYYY-MM-DD`
- Filled when task is moved to "Done"
- **`detail`**: Path to detail file
- Format: `./tasks/T-XXX.md`
- Example: `./tasks/T-001.md`
- Points to a markdown file with detailed description
- **`defaultExpanded`**: Whether task should be expanded by default
- Values: `true` or `false`
- When `true`, task shows all details when loaded
### Inline Description
If not using a detail file (`detail`), you can include inline description:
```markdown
### Task Name
- id: T-001
- tags: [backend]
- priority: high
```md
Detailed task description here.
Can contain multiple lines and markdown formatting.
```
```
## Detail Files (tasks/T-XXX.md)
When using `detail: ./tasks/T-XXX.md`, the file must follow this format:
```markdown
# T-001
- steps:
- [ ] Step 1 (not completed)
- [x] Step 2 (completed)
- [ ] Step 3 (not completed)
```md
Detailed task description.
Can include context, requirements, examples, etc.
```
```
### Steps Format
- Use `- [ ]` for uncompleted steps
- Use `- [x]` for completed steps
- Indentation: minimum 6 spaces before `-`
- Example:
```markdown
- steps:
- [ ] Research existing patterns
- [x] Create schema design
- [ ] Implement endpoints
```
## Important Rules
1. **Indentation**: All task properties must have 2 spaces indentation
2. **Date Format**: Always use `YYYY-MM-DD` for all dates
3. **Unique IDs**: Each task must have a unique ID in format `T-XXX`
4. **Tags**: Use arrays in format `[tag1, tag2]` with commas and spaces
5. **Descriptions**: Use markdown code blocks with ` ```md ` for descriptions
6. **Steps**: Only in detail files, not in main TASKS.md
7. **Sections**: Maintain Kanban sections (Backlog, Doing, Review, Done, Paused)
## Status Flow
- **Backlog** → **Doing**: When work starts
- **Doing** → **Review**: When awaiting review
- **Review** → **Done**: When approved/completed
- **Doing** → **Paused**: When temporarily paused
- **Paused** → **Doing**: When resumed
## Complete Examples
### Simple Task (no detail file)
```markdown
### Implement user authentication
- id: T-010
- tags: [authentication, security, backend, frontend]
- priority: high
- workload: Hard
- milestone: sprint-26-1_2
- start: 2026-01-15
- due: 2026-02-15
```
### Task with Inline Description
```markdown
### Create analytics dashboard
- id: T-011
- tags: [frontend, dashboard, analytics, visualization]
- priority: medium
- milestone: sprint-26-1_2
- start: 2026-01-15
- due: 2026-02-10
```md
Create an analytics visualization with charts and metrics.
Include filters by date and category.
```
```
### Task with Detail File
```markdown
### Implement n:n reconciliation between multiple sources
- id: T-001
- tags: [reconciliation, multiple-sources, n-to-n, backend, database]
- priority: high
- workload: Hard
- milestone: sprint-26-1_1
- start: 2026-01-01
- due: 2026-01-26
- detail: ./tasks/T-001.md
- defaultExpanded: false
```
## Common Mistakes to Avoid
1. Don't use incorrect indentation (must be 2 spaces for properties)
2. Don't mix date formats (always `YYYY-MM-DD`)
3. Don't forget to update `updated` when modifying tasks
4. Don't put steps directly in TASKS.md (only in detail files)
5. Don't use duplicate IDs
6. Don't forget to move tasks between sections when status changes
## Best Practices
1. Use sequential IDs (`T-001`, `T-002`, etc.)
2. Keep tags consistent and descriptive
3. Update `updated` whenever you modify a task
4. Use `detail` for complex tasks with many steps
5. Set `defaultExpanded: true` for important tasks
6. Use milestones to group related tasks
7. Keep Kanban sections organized
## References
- See `example-tasks/TASKS.md` for complete examples
- See `example-tasks/tasks/` for detail file examples
- See `example-tasks/README.md` for detailed documentation
Note: This ensures AI assistants understand the task format and maintain compatibility with the extension.
From VS Code Marketplace
Open VS Code Extensions view (Ctrl+Shift+X / Cmd+Shift+X)
tasks/T-010.md# T-010
- steps:
- [ ] Set up JWT token generation
- [ ] Implement login endpoint
- [ ] Create registration flow
- [ ] Add password reset functionality
```md
Implement complete authentication system with JWT tokens.
Include login, registration, and password reset functionality.
Requirements:
- Secure token storage
- Refresh token mechanism
- Password hashing with bcrypt
```
Important: Use exactly 2 spaces for indentation. All task properties must be indented with 2 spaces.
Features
Kanban Board View
Roadmap View
How to Open Kanban
Feature Overview
Kanban Board
Interactive Kanban board with drag & drop support
Roadmap View
Timeline view of tasks with start and due dates
Tag Filtering
Filter tasks by tags (supports multiple tags)
Sorting
Sort by name, date, priority, workload
Task Details
Support for detailed task files and inline descriptions
Real-time Sync
Two-way synchronization between Kanban and Markdown
AI Development Rules
For AI Assistants: This extension includes comprehensive rules for AI assistants to properly create and manage tasks. These rules ensure compatibility and maintain consistency across projects.
Rules Documentation
The project includes detailed rules in the rules/ directory:
Workflow: Maintain a tasks/ or kanban/ branch for task management. Feature branches reference tasks by ID. Periodically merge the tasks branch into main.
Pattern 4: Hybrid (Recommended)
Ideal for: Most projects, balance between simplicity and organization
Workflow: Main TASKS.md in main branch with overview. Feature branches reference tasks by ID in commits. Use milestones to group related tasks. For future planning, use a separate file (e.g., TASKS-PLANNING.md).
Pattern 5: TASKS.md per Feature/Epic
Ideal for: Projects with large features, agile methodology with epics