GitHub Spec Kit 5-phase workflow. Use when following the Constitution → Specify → Plan → Tasks → Implement cycle. Provides phase guidance, file templates, and workflow orchestration.
Implements the GitHub Spec Kit 5-phase workflow for specification-driven development. Use this to orchestrate the Constitution → Specify → Plan → Tasks → Implement cycle with structured templates and phase guidance.
/plugin marketplace add melodic-software/claude-code-plugins/plugin install spec-driven-development@melodic-softwareThis skill is limited to using the following tools:
references/constitution.mdreferences/phase-reference.mdreferences/prompts-guide.mdGitHub Spec Kit 5-phase specification-driven development workflow.
Keywords: Spec Kit, 5-phase workflow, constitution, specify, plan, tasks, implement, feature.md, design.md, tasks.md, specification workflow, GitHub Spec Kit, constitution.md, phased development
Use this skill when:
Phase 0: Constitution (One-time setup)
↓
Phase 1: Specify (Generate feature.md)
↓
Phase 2: Plan (Generate design.md)
↓
Phase 3: Tasks (Generate tasks.md)
↓
Phase 4: Implement (Code with guidance)
Establish project-wide principles, constraints, and non-functional requirements that guide ALL specifications.
.constitution.md# Project Constitution
## Core Principles
- <principle 1>
- <principle 2>
## Technical Constraints
- <constraint 1>
- <constraint 2>
## Quality Standards
- <standard 1>
- <standard 2>
## Non-Functional Requirements
- Performance: <requirement>
- Security: <requirement>
- Accessibility: <requirement>
Update the constitution when:
Transform a feature request or user story into a structured specification with EARS requirements and acceptance criteria.
feature.md# Feature: <Feature Name>
## Context
### Problem Statement
<What problem does this solve?>
### Motivation
<Why is this important?>
### Scope
<What's in/out of scope?>
## Requirements
### REQ-001: <Requirement Title>
**EARS Pattern:** <ubiquitous|state-driven|event-driven|unwanted|optional>
**Priority:** <must|should|could>
**Text:** <EARS-formatted requirement>
#### Acceptance Criteria
- **AC-001:**
- Given: <precondition>
- When: <action>
- Then: <outcome>
### REQ-002: ...
## Dependencies
- <dependency 1>
- <dependency 2>
## Risks
- <risk 1>
- <risk 2>
Design the technical implementation approach for the specified feature.
feature.md from Phase 1design.md# Design: <Feature Name>
## Overview
<High-level approach>
## Architecture
### Component Design
- <component 1>: <purpose>
- <component 2>: <purpose>
### Data Model
<Entities, relationships, schema changes>
### API Design
<Endpoints, contracts, interfaces>
## Technical Approach
### Approach Selected
<Chosen approach with rationale>
### Alternatives Considered
| Alternative | Pros | Cons | Why Not |
| --- | --- | --- | --- |
| <alt 1> | ... | ... | ... |
## Integration Points
- <integration 1>
- <integration 2>
## Testing Strategy
- Unit tests: <approach>
- Integration tests: <approach>
- E2E tests: <approach>
## Rollout Plan
<How will this be deployed?>
Break down the design into implementable tasks with clear deliverables.
design.md from Phase 2feature.md requirements for traceabilitytasks.md# Tasks: <Feature Name>
## Task List
### TSK-001: <Task Title>
**Status:** pending
**Requirement:** REQ-001
**Estimated Effort:** <S|M|L|XL>
**Description:**
<What needs to be done>
**Deliverables:**
- [ ] <deliverable 1>
- [ ] <deliverable 2>
**Acceptance Criteria:**
- [ ] <criterion 1>
- [ ] <criterion 2>
### TSK-002: ...
## Dependency Graph
```text
TSK-001 → TSK-003
TSK-002 → TSK-003
TSK-003 → TSK-004
```
## Effort Summary
| Size | Count | Typical Duration |
| --- | --- | --- |
| S | <n> | < 2 hours |
| M | <n> | 2-4 hours |
| L | <n> | 4-8 hours |
| XL | <n> | > 8 hours |
Execute tasks with continuous validation against specifications.
tasks.md from Phase 3design.md for technical guidancefeature.md for acceptance criteria1. Select next task (respecting dependencies)
2. Review task deliverables and acceptance criteria
3. Implement the task
4. Validate against acceptance criteria
5. Mark task complete
6. Repeat until all tasks done
Gate: Constitution exists and is current
Validation:
.constitution.md file existsGate: Specification is complete and valid
Validation:
Gate: Design is implementable
Validation:
Gate: Tasks are ready for implementation
Validation:
Located at: prompts/specify.prompt.md
Key sections:
Located at: prompts/plan.prompt.md
Key sections:
Located at: prompts/tasks.prompt.md
Key sections:
.specs/
├── <feature-name>/
│ ├── feature.md # Phase 1 output
│ ├── design.md # Phase 2 output
│ └── tasks.md # Phase 3 output
└── ...
.constitution.md # Phase 0 (project root)
| Phase | Command | Description |
|---|---|---|
| 0 | /spec:constitution | Create/update constitution |
| 1 | /spec:specify | Generate specification |
| 2 | /spec:plan | Generate design |
| 3 | /spec:tasks | Generate task breakdown |
| 4 | /spec:implement | Guided implementation |
| All | /spec:speckit:run | Full workflow |
Spec Kit outputs map to the canonical specification model:
| Spec Kit | Canonical Field |
|---|---|
| feature.md context | context.problem, context.motivation |
| REQ-xxx | requirements[].id |
| EARS text | requirements[].text |
| Priority | requirements[].priority |
| AC-xxx | requirements[].acceptance_criteria[] |
| design.md | implementation_notes |
Detailed Documentation:
Related Skills:
spec-management - Specification workflow navigationcanonical-spec-format - Canonical specification structureears-authoring - EARS requirement patternsgherkin-authoring - Acceptance criteria syntaxLast Updated: 2025-12-24
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.