# SpecLinter MCP - AI-Powered Specification Analysis
SpecLinter is a Model Context Protocol (MCP) server that transforms specifications into structured tasks with built-in quality gates for AI-powered development.
## Quick Setup
To use SpecLinter as an MCP server, add this to your AI IDE's MCP configuration:
```json
{
"mcpServers": {
"speclinter": {
"command": "node",
"args": ["/path/to/speclinter-mcp/dist/cli.js", "serve"],
"cwd": "/path/to/speclinter-mcp"
}
}
}
```
## Available MCP Tools
### Project Setup
- `speclinter_init_project` - Initialize SpecLinter in a project with smart defaults
### Codebase Analysis
- `speclinter_analyze_codebase` - Comprehensive codebase analysis that generates rich project documentation and context files
### Specification Parsing
- `speclinter_parse_spec` - Process specification and create SpecLinter tasks with AI analysis
### Similarity Detection
- `speclinter_find_similar` - Find similar features using AI-powered semantic analysis
### Task Management
- `speclinter_get_task_status` - Get current status of feature tasks
- `speclinter_update_task_status` - Update task status (completed, in-progress, blocked)
### Implementation Validation
- `speclinter_validate_implementation` - Scan codebase for feature implementation and provide comprehensive assessment
### AI-Powered Gherkin Generation
- `speclinter_generate_gherkin` - Generate comprehensive Gherkin scenarios for specific tasks with AI analysis
### Specification Quality Analysis
- `speclinter_analyze_spec_quality` - AI-powered specification quality analysis with semantic understanding
### Task Generation
- `speclinter_generate_tasks` - Generate comprehensive task breakdown from specifications with AI guidance
### Comprehensive Specification Analysis
- `speclinter_analyze_spec_comprehensive` - Complete specification analysis combining quality assessment and task generation
## Usage Examples
### Initialize Project
```
"Initialize SpecLinter in my project"
```
### Parse Specification
```
"Parse this spec: Create a user authentication system with email verification"
```
### Analyze Codebase
```
"Analyze my codebase and generate project documentation"
```
### Find Similar Features
```
"Check if this user profile feature is similar to existing features"
```
### Check Task Status
```
"Show me the status of my authentication-system tasks"
```
### Validate Implementation
```
"Validate the implementation of my authentication feature"
```
### Generate Gherkin Scenarios
```
"Generate Gherkin scenarios for my login task"
```
### Analyze Specification Quality
```
"Analyze the quality of this specification and provide improvement suggestions"
```
## Key Features
- **AI-Leveraged Analysis**: Semantic understanding vs regex patterns
- **Quality Grading**: A+ to F grading with improvement suggestions
- **Task Generation**: Break specs into implementable tasks with dependencies
- **AI-Powered Gherkin Scenarios**: Generate comprehensive, actionable test scenarios with multiple types (happy path, error handling, edge cases, validation)
- **Similarity Detection**: AI-powered semantic analysis to prevent duplicate work
- **Implementation Validation**: AI assessment of code against specifications
- **Quality Configuration**: Configurable scenario complexity, security/performance testing, and validation requirements
## Project Structure
SpecLinter creates this structure when initialized:
```
your-project/
├── .speclinter/
│ ├── config.json # Configuration
│ ├── speclinter.db # SQLite database
│ └── context/ # Project context
│ ├── project.md # Tech stack, constraints
│ ├── patterns.md # Code patterns
│ └── architecture.md # Architecture decisions
└── speclinter-tasks/
└── [feature-name]/
├── _active.md # Live status dashboard
├── task_01_*.md # Individual task files
├── meta.json # Feature metadata
└── gherkin/ # Test scenarios
└── *.feature
```
## Quality Grading System
- **A+ (95-100)**: Exceptional specification with comprehensive details
- **A (90-94)**: Excellent specification with clear requirements
- **B (80-89)**: Good specification with minor improvements needed
- **C (70-79)**: Adequate specification with several areas for improvement
- **D (60-69)**: Poor specification requiring significant enhancement
- **F (0-59)**: Failing specification with major issues
## Installation
```bash
# Install dependencies and build
pnpm install && pnpm build
# Alternative: use npm or yarn
npm install && npm run build
```
## Requirements
- Node.js 18+
- TypeScript support
- SQLite for task management
## Repository
https://github.com/orangebread/speclinter-mcp
## License
MIT
