Igniral MCP Server
Bridges AI agents with Igniral's platform to generate production-ready backends from natural language descriptions.
README
Igniral MCP Server
Igniral — Production-Ready Backends with AI Speed
Model Context Protocol (MCP) server that bridges AI agents (Claude, Cursor, Antigravity, etc.) with Igniral's backend platform. Describe your API in plain English — Igniral generates the schema, CRUD endpoints, authentication, Swagger docs, and antivirus-protected file storage automatically.
What is Igniral?
Igniral lets you generate 100% of your API infrastructure with a simple prompt, or build manually using a Visual Schema Builder. Everything is production-ready from the start:
- 🤖 AI-Powered Generation — Describe your data model, get a complete REST API instantly
- 🔐 Built-in Auth & RBAC — JWT authentication with role-based access control, no auth code needed
- 📄 Always-Sync Swagger — OpenAPI docs update automatically with every change
- 🛡️ Antivirus File Storage — Every uploaded file is scanned by ClamAV before reaching your infrastructure
- 📊 Real-time Analytics — Monitor API usage, error rates, and traffic from your dashboard
- 🗄️ Managed Database — Automatic backups and replication, zero DBA required
Get started for free at igniral.com → Start Now
Prerequisites
- Create an Igniral account at igniral.com (free tier available)
- Generate Agent API Keys from the Igniral Dashboard → Agent API Keys
- Have Node.js ≥ 18 installed on your machine
Important: This is an MCP server — it runs inside your AI-powered IDE (Claude Desktop, Cursor, Antigravity, etc.), not directly from the terminal. You configure it once in your IDE settings, and the IDE handles starting and stopping it automatically.
Quick Start
Choose your IDE and add the following configuration. Replace agent-xxxxxxxxxxxx and your-client-secret with your actual Agent API Key credentials.
Claude Desktop
Edit your claude_desktop_config.json:
{
"mcpServers": {
"igniral": {
"command": "npx",
"args": ["-y", "igniral-mcp-server"],
"env": {
"IGNIRAL_CLIENT_ID": "agent-xxxxxxxxxxxx",
"IGNIRAL_CLIENT_SECRET": "your-client-secret"
}
}
}
}
Cursor
Edit your .cursor/mcp.json:
{
"mcpServers": {
"igniral": {
"command": "npx",
"args": ["-y", "igniral-mcp-server"],
"env": {
"IGNIRAL_CLIENT_ID": "agent-xxxxxxxxxxxx",
"IGNIRAL_CLIENT_SECRET": "your-client-secret"
}
}
}
}
Antigravity (Google)
Edit ~/.gemini/antigravity/mcp_config.json:
{
"mcpServers": {
"igniral": {
"command": "npx",
"args": ["-y", "igniral-mcp-server"],
"env": {
"IGNIRAL_CLIENT_ID": "agent-xxxxxxxxxxxx",
"IGNIRAL_CLIENT_SECRET": "your-client-secret"
}
}
}
}
Note: Antigravity may not inherit your shell's
PATH. Use the absolute path tonode(e.g.,/opt/homebrew/Cellar/node/25.9.0_2/bin/node) if you get "executable not found" errors.
That's it! After saving the configuration and restarting your IDE, you can ask your AI agent things like:
- "Build me a gym management API"
- "Create a REST API for a pet store with products, orders, and users"
- "List my existing Igniral applications"
Tools
Once configured, the following tools are available to your AI agent:
| Tool | Description |
|---|---|
igniral_generate_schema_from_prompt |
Auto-generate a complete app from a natural language description |
igniral_create_application |
Create an empty application shell manually |
igniral_create_dynamic_endpoint |
Add API endpoints to an existing application |
igniral_list_applications |
List the user's existing applications |
Architecture
AI Agent (Claude/Cursor)
│
▼ (MCP Protocol - stdio)
┌──────────────────────────┐
│ Igniral MCP Server │
│ ├─ TokenManager │ ← OAuth2 client_credentials
│ ├─ Zod Validation │
│ ├─ SSE Client │
│ └─ HTTP Client │
└────────────┬─────────────┘
│ (HTTP + JWT with sub=userId)
┌───────┼───────┐
▼ ▼ ▼
auth- ai-schema- json-elements
server builder microservice
(token) (auto-gen) (CRUD)
Environment Variables
| Variable | Required | Description |
|---|---|---|
IGNIRAL_CLIENT_ID |
✅ | Agent API Key client ID (from Dashboard) |
IGNIRAL_CLIENT_SECRET |
✅ | Agent API Key client secret (shown once at creation) |
IGNIRAL_AUTH_URL |
❌ | Auth server URL (default: https://auth.igniral.com) |
IGNIRAL_API_URL |
❌ | API URL (default: https://api.igniral.io) |
IGNIRAL_AI_API_URL |
❌ | AI API URL (default: https://ai.igniral.com) |
Development (from source)
Only needed if you want to contribute or modify the server:
git clone https://github.com/igniral/igniral-mcp-server.git
cd igniral-mcp-server
npm install
cp .env.example .env # Edit with your credentials
npm run dev # Start in development mode
npm run inspect # Test with MCP Inspector
Links
- 🌐 Website: igniral.com
- 📖 Documentation: igniral.com/docs
- 💬 Community: Discord
- 📧 Support: support@igniral.com
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。