kiro-spec-tasks

Generate implementation tasks from feature specs into tasks.md with dependency annotations.

3.6k|274|Updated Jul 17, 2025
One-click install
npx skills add https://github.com/gotalab/cc-sdd --skill kiro-spec-tasks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kiro-spec-tasks
Source: https://github.com/gotalab/cc-sdd/tree/main/tools/cc-sdd/templates/agents/claude-code-skills/skills/kiro-spec-tasks
Command: npx skills add https://github.com/gotalab/cc-sdd --skill kiro-spec-tasks

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill converts approved requirements and design artifacts into an executable, reviewable tasks.md that maps numeric requirement IDs to concrete 1-3 hour implementation tasks while enforcing boundaries, dependencies, and parallelization rules.

Core Features & Use Cases

  • Requirement-to-task mapping: Ensures every numeric requirement ID appears in at least one task and that tasks include observable done criteria.
  • Sizing, boundaries, and dependencies: Sizes tasks to 1-3 hours, annotates architecture boundaries, and records explicit cross-boundary dependencies and parallel (P) markers when safe.
  • Quality gates and outputs: Runs a task-plan review gate and an independent task-graph sanity review, merges with existing tasks.md when present, writes tasks.md, and updates spec.json metadata and approvals; supports auto-approval flags for unattended runs.
  • Use Case: Prepare a feature spec for autonomous implementation by producing a validated task graph and ready-to-run implementation plan.

Quick Start

Run the kiro-spec-tasks skill for the desired feature to generate tasks.md, perform the task-plan sanity checks, and update spec.json accordingly.

Frequently Asked Questions about kiro-spec-tasks

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate implementation tasks from requirements for spec-driven development?

To generate implementation tasks from requirements, you map approved requirement IDs and design artifacts to concrete 1-3 hour work units, producing an executable tasks.md file with defined boundaries and dependencies.

What is the best way to size tasks for autonomous implementation planning?

The best way to size tasks for autonomous implementation planning is to constrain each work unit to 1-3 hours, annotate architecture boundaries, and mark parallelizable tasks with explicit cross-boundary dependencies.

How does a task-graph sanity check work during task generation?

A task-graph sanity check works by running an independent review alongside a task-plan review gate to validate that every numeric requirement ID maps to at least one task with observable done criteria before updating spec.json.

Can I merge newly generated tasks with an existing tasks.md file?

Yes, you can merge newly generated tasks with an existing tasks.md file. The generation process merges new task definitions with existing content, writes the updated tasks.md, and updates spec.json metadata and approvals.

Does task generation support auto-approval for unattended spec-driven runs?

Yes, task generation supports auto-approval flags for unattended runs. This allows the task-plan review and task-graph sanity check to execute and update spec.json metadata without manual intervention during spec-driven development.

When do I need to map numeric requirement IDs to implementation tasks?

You need to map numeric requirement IDs to implementation tasks when preparing a feature spec for autonomous implementation, ensuring every requirement has corresponding tasks with observable done criteria and validated dependency graphs.