Files
claude-task-master/.kiro/steering/taskmaster_hooks_workflow.md
Ralph Khreish 4423119a5e feat: Add Kiro hooks and configuration for Taskmaster integration (#1032)
* feat: Add Kiro hooks and configuration for Taskmaster integration

- Introduced multiple Kiro hooks to automate task management workflows, including:
  - Code Change Task Tracker
  - Complexity Analyzer
  - Daily Standup Assistant
  - Git Commit Task Linker
  - Import Cleanup on Delete
  - New File Boilerplate
  - PR Readiness Checker
  - Task Dependency Auto-Progression
  - Test Success Task Completer
- Added .mcp.json configuration for Taskmaster AI integration.
- Updated development workflow documentation to reflect new hook-driven processes and best practices.

This commit enhances the automation capabilities of Taskmaster, streamlining task management and improving developer efficiency.

* chore: run format

* chore: improve unit tests on kiro rules

* chore: run format

* chore: run format

* feat: improve PR and add changeset
2025-07-23 17:02:16 +02:00

2.1 KiB

inclusion
inclusion
always

Taskmaster Hook-Driven Workflow

Core Principle: Hooks Automate Task Management

When working with Taskmaster in Kiro, avoid manually marking tasks as done. The hook system automatically handles task completion based on:

  • Test Success: [TM] Test Success Task Completer detects passing tests and prompts for task completion
  • Code Changes: [TM] Code Change Task Tracker monitors implementation progress
  • Dependency Chains: [TM] Task Dependency Auto-Progression auto-starts dependent tasks

AI Assistant Workflow

Follow this pattern when implementing features:

  1. Implement First: Write code, create tests, make changes
  2. Save Frequently: Hooks trigger on file saves to track progress automatically
  3. Let Hooks Decide: Allow hooks to detect completion rather than manually setting status
  4. Respond to Prompts: Confirm when hooks suggest task completion

Key Rules for AI Assistants

  • Never use tm set-status --status=done unless hooks fail to detect completion
  • Always write tests - they provide the most reliable completion signal
  • Save files after implementation - this triggers progress tracking
  • Trust hook suggestions - if no completion prompt appears, more work may be needed

Automatic Behaviors

The hook system provides:

  • Progress Logging: Implementation details automatically added to task notes
  • Evidence-Based Completion: Tasks marked done only when criteria are met
  • Dependency Management: Next tasks auto-started when dependencies complete
  • Natural Flow: Focus on coding, not task management overhead

Manual Override Cases

Only manually set task status for:

  • Documentation-only tasks
  • Tasks without testable outcomes
  • Emergency fixes without proper test coverage

Use tm set-status sparingly - prefer hook-driven completion.

Implementation Pattern

1. Implement feature → Save file
2. Write tests → Save test file
3. Tests pass → Hook prompts completion
4. Confirm completion → Next task auto-starts

This workflow ensures proper task tracking while maintaining development flow.