speckitmcp
MCP server that integrates GitHub Spec-Kit with AI coding agents to manage Spec-Driven Development workflows, including specification authoring, planning, task generation, and consistency analysis.
README
<p align="center"> <h1 align="center">speckitmcp</h1> <p align="center"> A <a href="https://modelcontextprotocol.io">Model Context Protocol</a> server for <a href="https://github.com/github/spec-kit">GitHub Spec-Kit</a><br/> <em>Bring Spec-Driven Development to any AI coding agent.</em> </p> </p>
<p align="center"> <a href="#quick-start">Quick Start</a> • <a href="#tools">Tools</a> • <a href="#resources">Resources</a> • <a href="#prompts">Prompts</a> • <a href="#the-sdd-workflow">Workflow</a> • <a href="#development">Development</a> </p>
Overview
speckitmcp bridges GitHub's Spec-Kit toolkit with any MCP-compatible AI assistant — Claude Code, Cursor, VS Code Copilot, Windsurf, and more.
It exposes the full Spec-Driven Development (SDD) workflow as MCP tools, resources, and prompts, so your AI agent can:
- Initialize and manage spec-kit projects
- Author specifications, technical plans, and task breakdowns
- Track implementation progress and mark tasks complete
- Validate cross-artifact consistency with a 6-pass analysis engine
- Generate quality checklists and convert tasks to GitHub issues
All without leaving your editor.
Quick Start
Prerequisites
| Requirement | Install |
|---|---|
| Node.js 18+ | nodejs.org |
| spec-kit CLI | uv tool install --from git+https://github.com/github/spec-kit.git specify-cli |
Install & Build
git clone https://github.com/jthom233/speckitmcp.git
cd speckitmcp
npm install
npm run build
Connect to Your AI Agent
<details> <summary><strong>Claude Code</strong></summary>
Add to ~/.claude/settings.json (global) or .claude/settings.json (project):
{
"mcpServers": {
"spec-kit": {
"command": "node",
"args": ["/absolute/path/to/speckitmcp/dist/index.js"]
}
}
}
</details>
<details> <summary><strong>Cursor</strong></summary>
Add to .cursor/mcp.json in your project root:
{
"mcpServers": {
"spec-kit": {
"command": "node",
"args": ["/absolute/path/to/speckitmcp/dist/index.js"]
}
}
}
</details>
<details> <summary><strong>VS Code / GitHub Copilot</strong></summary>
Add to .vscode/mcp.json in your project root:
{
"servers": {
"spec-kit": {
"command": "node",
"args": ["/absolute/path/to/speckitmcp/dist/index.js"]
}
}
}
</details>
<details> <summary><strong>Windsurf / Other MCP Clients</strong></summary>
Point your client's MCP configuration at:
node /absolute/path/to/speckitmcp/dist/index.js
The server communicates over stdio using the standard MCP JSON-RPC protocol. </details>
Tools
The server exposes 13 tools that map to every phase of the SDD workflow:
| Tool | Description |
|---|---|
speckit_init |
Initialize spec-kit project structure |
speckit_check |
Check spec-kit installation and system prerequisites |
speckit_version |
Get spec-kit CLI version |
speckit_status |
View project status with task completion stats |
speckit_constitution |
Read, create, or update project constitution with optional version bumping |
speckit_specify |
Create feature spec with template loading, script integration, and auto-generated quality checklist |
speckit_plan |
Multi-phase planning (research, design, plan) with spec prerequisite and constitution gate |
speckit_tasks |
Generate user-story-organized task lists (requires spec.md and plan.md) |
speckit_implement |
Read tasks/docs, mark tasks complete with checklist gate, or add notes |
speckit_clarify |
Scan spec.md for ambiguities (9 categories, max 5 questions) or answer them inline |
speckit_analyze |
Read-only 6-pass analysis: duplication, ambiguity, underspecification, constitution alignment, coverage gaps, inconsistency |
speckit_checklist |
Generate requirement quality checklists (spec quality, not implementation) in checklists/ subdirectory |
speckit_tasks_to_issues |
Convert tasks.md to GitHub issues (dry-run by default) |
Key Features
- Script integration — Platform-aware bash/powershell helper scripts invoked automatically
- Template loading — Loads templates from
.specify/templates/with embedded fallbacks - Prerequisites checking — Tools verify prior artifacts exist before proceeding
- Constitution gate — Planning phase reads the project constitution as mandatory context
- Checklist gate — Implementation checks for incomplete checklist items before marking tasks done
- 6-pass analysis engine — Catches duplication, ambiguity, underspecification, constitution violations, coverage gaps, and inconsistencies
- GitHub issue creation — Convert a tasks.md file into GitHub issues via
speckit_tasks_to_issues
Resources
The server exposes spec-kit project files as read-only MCP resources:
| URI | Content |
|---|---|
speckit://constitution |
Project constitution |
speckit://templates/{name} |
Spec-kit templates |
speckit://specs/{feature}/spec |
Feature specification |
speckit://specs/{feature}/plan |
Implementation plan |
speckit://specs/{feature}/tasks |
Task list |
speckit://specs/{feature}/research |
Research notes |
speckit://specs/{feature}/data-model |
Data model |
speckit://specs/{feature}/quickstart |
Quickstart guide |
speckit://specs/{feature}/checklists/{name} |
Quality checklists |
speckit://specs/{feature}/contracts/{name} |
API contracts |
Prompts
Ten built-in prompts guide your AI agent through each SDD phase:
| Prompt | Purpose |
|---|---|
sdd_workflow |
End-to-end walkthrough of the full SDD lifecycle |
sdd_specify |
Structured feature specification authoring |
sdd_clarify |
Guided ambiguity resolution |
sdd_plan |
Technical planning with architecture and stack decisions |
sdd_tasks |
Task breakdown with phasing, dependencies, and parallelism |
sdd_implement |
Task execution and progress tracking |
sdd_checklist |
Requirement quality checklist generation |
sdd_analyze |
Cross-artifact consistency validation |
sdd_constitution |
Guided creation of project principles and governance |
sdd_taskstoissues |
Convert tasks to GitHub issues |
The SDD Workflow
init → constitution → specify → clarify → plan → tasks → checklist → analyze → implement → tasks_to_issues
- Init —
speckit_init— Scaffold the project with.specify/templates - Constitution —
speckit_constitution— Define project principles and governance before specifying - Specify —
speckit_specify— Define requirements as prioritized user stories with template loading - Clarify —
speckit_clarify— Scan for and resolve ambiguities before committing to a plan - Plan —
speckit_plan— Multi-phase planning gated on project constitution - Tasks —
speckit_tasks— Generate user-story-organized task lists with prerequisites check - Checklist —
speckit_checklist— Generate requirement quality checklists - Analyze —
speckit_analyze— 6-pass validation of spec / plan / tasks alignment - Implement —
speckit_implement— Track completion with checklist gate and regex-safe marking - Tasks to Issues —
speckit_tasks_to_issues— Push tasks to GitHub as issues
Project Structure
src/
├── index.ts # Entry point — stdio transport
├── server.ts # MCP server — handler registration
├── cli.ts # spec-kit CLI wrapper (child_process)
├── tools/
│ ├── index.ts # Tool registry
│ ├── init.ts # speckit_init
│ ├── check.ts # speckit_check
│ ├── version.ts # speckit_version
│ ├── status.ts # speckit_status
│ ├── constitution.ts # speckit_constitution
│ ├── specify.ts # speckit_specify
│ ├── plan.ts # speckit_plan
│ ├── tasks.ts # speckit_tasks
│ ├── implement.ts # speckit_implement
│ ├── clarify.ts # speckit_clarify
│ ├── analyze.ts # speckit_analyze
│ ├── checklist.ts # speckit_checklist
│ └── tasks-to-issues.ts # speckit_tasks_to_issues
├── resources/
│ └── index.ts # MCP resource handlers
└── prompts/
└── index.ts # SDD workflow prompts
Development
npm install # Install dependencies
npm run build # Compile TypeScript → dist/
npm run dev # Watch mode (rebuild on change)
Manual Smoke Test
echo '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"test","version":"1.0"}}}' | node dist/index.js
You should see a JSON-RPC response with serverInfo.name: "spec-kit-mcp".
Tech Stack
| Layer | Choice |
|---|---|
| Language | TypeScript 5 (strict mode, ESM) |
| Runtime | Node.js 18+ |
| Protocol | @modelcontextprotocol/sdk v1.x |
| Validation | Zod |
| Transport | stdio (JSON-RPC 2.0) |
| CLI Integration | Node.js child_process wrapping specify |
Contributing
Contributions are welcome! Please open an issue first to discuss what you'd like to change.
- Fork the repo
- Create a feature branch (
git checkout -b feat/my-feature) - Commit your changes
- Push to your fork and open a Pull Request
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.