design-documentation

安装量: 60
排名: #12475

安装

npx skills add https://github.com/jasonkneen/kiro --skill design-documentation

Design Documentation

Create technical blueprints that bridge requirements and implementation. This skill teaches how to document architecture decisions, component design, and system interactions.

When to Use This Skill

Use design documentation when:

Requirements phase is complete and approved You need to plan technical implementation Multiple developers will work on the feature Architecture decisions need documentation The feature involves complex integrations Design Document Structure Standard Template

Design Document: [Feature Name]

Overview

[High-level summary of the feature and approach]

Architecture

[System architecture and component overview]

Components and Interfaces

[Detailed component descriptions and interactions]

Data Models

[Data structures and relationships]

Error Handling

[Error scenarios and response strategies]

Testing Strategy

[Testing approach and quality assurance]

Step-by-Step Process Step 1: Requirements Analysis

Before designing, ensure you understand:

All functional requirements Non-functional requirements (performance, security, scalability) Constraints (technology stack, timeline, resources) Integration points with existing systems

Analysis Questions:

What does the system need to do? What are the performance expectations? What existing code/systems does this touch? What are the security requirements? What could go wrong? Step 2: Research and Context Building

Identify areas needing research:

Technology choices and alternatives Third-party integrations and APIs Best practices for similar systems Security and compliance considerations

Research Documentation:

Research Findings

[Topic]

Sources: [Links/references] Key Insights: - [Finding 1] - [Finding 2] Impact on Design: [How this affects our approach]

Step 3: Define System Architecture

Document the high-level structure:

Architecture

System Overview

[How the overall system works]

Component Architecture

[Major components and their responsibilities]

Data Flow

[How information moves through the system]

Technology Decisions

[Key technology choices and rationale]

Architecture Diagram (Mermaid):

graph TD A[Client] --> B[API Gateway] B --> C[Auth Service] B --> D[Feature Service] D --> E[Database] D --> F[Cache]

Step 4: Design Components and Interfaces

For each major component:

Components and Interfaces

[Component Name]

Purpose: [What this component does]

Responsibilities: - [Responsibility 1] - [Responsibility 2]

Interfaces: - Input: [What it receives] - Output: [What it produces] - Dependencies: [What it requires]

API Definition: ```typescript interface ComponentName { method(input: InputType): Promise; }

Step 5: Define Data Models

Document all data structures:

```markdown

Data Models

[Entity Name]

Purpose: [What this entity represents]

Properties: | Field | Type | Required | Description | |-------|------|----------|-------------| | id | string | Yes | Unique identifier | | name | string | Yes | Display name | | createdAt | Date | Yes | Creation timestamp |

Validation Rules: - [Rule 1] - [Rule 2]

Relationships: - [Relationship to other entities]

Example: ```json { "id": "abc123", "name": "Example", "createdAt": "2024-01-15T10:30:00Z" }

Step 6: Plan Error Handling

Document error scenarios:

```markdown

Error Handling

Error Categories

  1. Validation Errors: Invalid user input
  2. Authentication Errors: Unauthorized access
  3. External Service Errors: Third-party failures
  4. System Errors: Internal failures

Error Response Strategy

Error Type HTTP Code User Message System Action
Validation 400 Specific field error Log, return details
Auth 401 "Please log in" Redirect to login
Not Found 404 "Resource not found" Log, return error
Server 500 "Something went wrong" Log, alert, retry

Recovery Mechanisms

  • [Retry strategies]
  • [Fallback behaviors]
  • [Circuit breaker patterns]

Step 7: Define Testing Strategy

Testing Strategy

Unit Testing

  • Coverage Target: 80%+
  • Focus Areas: [Critical business logic]
  • Mocking Strategy: [What to mock]

Integration Testing

  • Scope: [Component interactions to test]
  • Environment: [Test environment setup]
  • Data Strategy: [Test data approach]

End-to-End Testing

  • Critical Paths: [User journeys to test]
  • Tools: [Testing tools]

Performance Testing

  • Load Targets: [Expected load]
  • Benchmarks: [Performance requirements]

Decision Documentation

Document key decisions using this template:

Decision: [Brief Title]

Context: [Situation requiring a decision]

Options Considered:

Option 1: [Name] - Pros: [Benefits] - Cons: [Drawbacks] - Effort: [Low/Medium/High]

Option 2: [Name] - Pros: [Benefits] - Cons: [Drawbacks] - Effort: [Low/Medium/High]

Decision: [Chosen option]

Rationale: [Why this option was selected]

Implications: [What this means for implementation]

Common Design Decisions API Design

Decision: API Architecture

Options: 1. REST API - Standard HTTP methods, well-understood 2. GraphQL - Flexible queries, single endpoint 3. RPC-style - Direct operation mapping

Decision: REST API Rationale: Standard CRUD operations, team familiarity, good tooling support

Data Storage

Decision: Database Choice

Options: 1. PostgreSQL - ACID compliance, complex queries 2. MongoDB - Schema flexibility, horizontal scaling 3. Redis - High performance, limited queries

Decision: PostgreSQL with JSON columns Rationale: Data consistency needs + flexibility for user preferences

Authentication

Decision: Authentication Strategy

Options: 1. Session-based - Server-controlled, simple 2. JWT tokens - Stateless, scalable 3. OAuth 2.0 - External provider, no password management

Decision: JWT with refresh token rotation Rationale: Scalability requirements, API-first architecture

Quality Checklist

Before finalizing design:

Completeness:

All requirements addressed in design Major system components defined Data models cover all entities Error handling covers expected failures Testing strategy addresses all layers

Clarity:

Design decisions clearly explained Component responsibilities well-defined Interfaces between components specified Technical choices include rationale

Feasibility:

Design is technically achievable Performance requirements can be met Security requirements addressed Implementation complexity reasonable

Traceability:

Design elements map to requirements All requirements covered by design Testing validates requirement fulfillment Common Pitfalls Over-Engineering: Design for current requirements, not hypothetical futures Under-Specified Interfaces: Define clear component boundaries Ignoring Non-Functional Requirements: Address performance, security, scalability Technology-First Design: Let requirements drive technology choices Insufficient Error Handling: Plan for failures, not just happy paths Example: User Profile Design

Design Document: User Profile Management

Overview

Enable users to view and edit their profile information, including personal details, preferences, and profile picture.

Architecture

Component Architecture

┌─────────────┐ ┌──────────────┐ ┌─────────────┐ │ Client │────▶│ Profile API │────▶│ Database │ └─────────────┘ └──────────────┘ └─────────────┘ │ ▼ ┌──────────────┐ │ Image Store │ └──────────────┘

Technology Stack

  • API: Node.js with Express
  • Database: PostgreSQL
  • Image Storage: S3-compatible object storage
  • Cache: Redis for session data

Components

ProfileService

Purpose: Manage user profile CRUD operations

Interface: ```typescript interface ProfileService { getProfile(userId: string): Promise; updateProfile(userId: string, data: ProfileUpdate): Promise; uploadAvatar(userId: string, image: File): Promise; deleteAvatar(userId: string): Promise; }

Data Models UserProfile Field Type Required Validation id UUID Yes Auto-generated email string Yes Valid email format displayName string Yes 2-50 characters bio string No Max 500 characters avatarUrl string No Valid URL preferences JSON No Valid JSON object updatedAt timestamp Yes Auto-updated Error Handling Scenario Response Action Profile not found 404 Return error message Invalid update data 400 Return validation errors Image too large 413 Return size limit error Unauthorized 401 Redirect to login Testing Strategy Unit tests for validation logic Integration tests for API endpoints E2E test for profile edit flow

Next Steps

After completing design: 1. Get design review and approval 2. Move to Task Planning phase 3. Break design into implementation tasks 4. Begin systematic implementation

返回排行榜