ServiceNow MCP Server (SSE)
A browser-accessible Model Context Protocol server that enables AI assistants to manage ServiceNow incidents using Server-Sent Events. It provides comprehensive tools for creating, listing, retrieving, and updating incidents through a type-safe RESTful interface.
README
ServiceNow MCP Server (SSE - Browser Accessible)
A browser-accessible Model Context Protocol (MCP) server for ServiceNow incident management using Server-Sent Events (SSE) transport.
🌟 Key Features
- ✅ Browser Accessible - Access via HTTP/SSE from any browser
- ✅ AI Assistant Compatible - Works with Bob, Claude, and other MCP clients
- ✅ Four ServiceNow Tools - Complete incident management (get, list, create, update)
- ✅ RESTful Health Checks - Monitor server status
- ✅ CORS Enabled - Cross-origin requests supported
- ✅ Type-Safe - Full TypeScript implementation with Zod validation
🆚 Difference from Stdio Version
| Feature | Stdio Server | SSE Server (This One) |
|---|---|---|
| Browser Access | ❌ No | ✅ Yes |
| AI Assistant Access | ✅ Yes | ✅ Yes |
| HTTP Endpoints | ❌ No | ✅ Yes |
| Port Required | ❌ No | ✅ Yes (3000) |
| Testing | Command line only | Browser, curl, Postman |
| Use Case | AI assistants only | Browsers + AI assistants |
📦 Installation
Prerequisites
- Node.js v18+ (you have v24.14.0 ✅)
- npm
- ServiceNow instance with API access
Setup
-
Install dependencies:
cd servicenow-server-sse /usr/local/bin/npm install -
Build the server:
/usr/local/bin/npm run build -
Set environment variables:
export SERVICENOW_INSTANCE_URL="https://your-instance.service-now.com" export SERVICENOW_USERNAME="your-username" export SERVICENOW_PASSWORD="your-password" export PORT=3000 # Optional, defaults to 3000 -
Start the server:
/usr/local/bin/npm start
🚀 Quick Start
Start the Server
# Set credentials
export SERVICENOW_INSTANCE_URL="https://dev12345.service-now.com"
export SERVICENOW_USERNAME="admin"
export SERVICENOW_PASSWORD="your-password"
# Start server
cd servicenow-server-sse
/usr/local/bin/npm start
You should see:
🚀 ServiceNow MCP Server (SSE) running on http://localhost:3000
📡 SSE endpoint: http://localhost:3000/sse
💚 Health check: http://localhost:3000/health
📖 API info: http://localhost:3000/
🔗 ServiceNow instance: https://dev12345.service-now.com
🌐 Browser Access
Health Check
Open in browser: http://localhost:3000/health
Response:
{
"status": "healthy",
"server": "servicenow-server-sse",
"version": "0.1.0",
"servicenow_instance": "https://your-instance.service-now.com"
}
API Information
Open in browser: http://localhost:3000/
Shows available tools and endpoints.
SSE Connection
The MCP protocol endpoint: http://localhost:3000/sse
This is used by MCP clients (AI assistants) to connect.
🛠️ Available Tools
1. get_incident
Retrieve incident details by number or sys_id.
Parameters:
incident_id(required): Incident number (e.g., "INC0010001") or sys_id
Example:
{
"incident_id": "INC0010001"
}
2. list_incidents
Query incidents with filters and pagination.
Parameters:
state(optional): 1=New, 2=In Progress, 3=On Hold, 6=Resolved, 7=Closedpriority(optional): 1=Critical, 2=High, 3=Moderate, 4=Low, 5=Planningassigned_to(optional): User sys_id or usernameassignment_group(optional): Group sys_id or namecaller_id(optional): Caller sys_id or usernamecreated_after(optional): ISO 8601 datecreated_before(optional): ISO 8601 datelimit(optional): 1-100, default 10offset(optional): Pagination offset, default 0
Example:
{
"state": "2",
"priority": "1",
"limit": 20
}
3. create_incident
Create a new incident.
Parameters:
short_description(required): Brief descriptioncaller_id(required): Caller sys_id or usernamedescription(optional): Detailed descriptionurgency(optional): 1=High, 2=Medium, 3=Lowimpact(optional): 1=High, 2=Medium, 3=Lowpriority(optional): 1-5assignment_group(optional): Group sys_id or nameassigned_to(optional): User sys_id or usernamecategory(optional): Categorysubcategory(optional): Subcategory
Example:
{
"short_description": "Email server down",
"caller_id": "john.doe",
"urgency": "1",
"impact": "1"
}
4. update_incident
Update an existing incident.
Parameters:
incident_id(required): Incident number or sys_idstate(optional): New statepriority(optional): New priorityassigned_to(optional): Assign to userassignment_group(optional): Assign to groupwork_notes(optional): Add work notesclose_notes(optional): Close notesresolution_code(optional): Resolution codeshort_description(optional): Update descriptiondescription(optional): Update detailed description
Example:
{
"incident_id": "INC0010001",
"state": "2",
"assigned_to": "jane.smith",
"work_notes": "Investigating the issue"
}
🔌 Using with AI Assistants
Configure Bob
Edit ~/.bob/settings/mcp_settings.json:
{
"mcpServers": {
"servicenow-sse": {
"url": "http://localhost:3000/sse",
"disabled": false,
"alwaysAllow": [],
"disabledTools": []
}
}
}
Note: For SSE servers, use url instead of command and args.
Configure Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"servicenow-sse": {
"url": "http://localhost:3000/sse"
}
}
}
Restart Your AI Assistant
After configuration, restart Bob or Claude Desktop to connect to the server.
🧪 Testing
Using curl
# Health check
curl http://localhost:3000/health
# API info
curl http://localhost:3000/
# SSE connection (will stream events)
curl -N http://localhost:3000/sse
Using Browser
- Open
http://localhost:3000/to see API information - Open
http://localhost:3000/healthto check server health - Use browser developer tools to test SSE connection
Using Postman
- Import the server URL:
http://localhost:3000 - Test health endpoint: GET
/health - Test SSE endpoint: GET
/sse(use EventSource)
🔒 Security Considerations
Production Deployment
For production use, consider:
- Authentication: Add API key or OAuth authentication
- HTTPS: Use TLS/SSL certificates
- Rate Limiting: Implement request rate limits
- CORS: Restrict allowed origins
- Environment Variables: Use secure secret management
- Firewall: Restrict access to trusted IPs
Example with Authentication
// Add to Express middleware
app.use((req, res, next) => {
const apiKey = req.headers['x-api-key'];
if (apiKey !== process.env.API_KEY) {
return res.status(401).json({ error: 'Unauthorized' });
}
next();
});
📊 Monitoring
Health Check Endpoint
curl http://localhost:3000/health
Returns server status and configuration.
Logs
The server logs to console:
- Connection events
- Request handling
- Errors and warnings
Process Management
Use PM2 for production:
npm install -g pm2
pm2 start build/index.js --name servicenow-sse
pm2 logs servicenow-sse
pm2 restart servicenow-sse
🐛 Troubleshooting
Port Already in Use
# Find process using port 3000
lsof -i :3000
# Kill the process
kill -9 <PID>
# Or use a different port
export PORT=3001
npm start
CORS Errors
The server has CORS enabled for all origins. If you need to restrict:
app.use(cors({
origin: 'https://your-domain.com',
methods: ['GET', 'POST'],
}));
ServiceNow Connection Issues
- Verify credentials are correct
- Check ServiceNow instance URL (no trailing slash)
- Ensure user has
itilrole - Test ServiceNow API directly with curl
SSE Connection Drops
- Check network stability
- Implement reconnection logic in client
- Monitor server logs for errors
🚀 Deployment
Local Development
npm run dev
Production
# Build
npm run build
# Start with PM2
pm2 start build/index.js --name servicenow-sse
# Or with systemd
sudo systemctl start servicenow-sse
Docker
FROM node:18-alpine
WORKDIR /app
COPY package*.json ./
RUN npm install
COPY . .
RUN npm run build
EXPOSE 3000
CMD ["node", "build/index.js"]
📚 API Endpoints
| Endpoint | Method | Description |
|---|---|---|
/ |
GET | API information and available tools |
/health |
GET | Health check and server status |
/sse |
GET | MCP SSE connection endpoint |
/message |
POST | MCP message handling (used by SSE) |
🔄 Comparison with Stdio Server
Both servers provide the same ServiceNow tools, but differ in access method:
Use Stdio Server when:
- Only using with AI assistants (Bob, Claude)
- Don't need browser access
- Want simpler configuration
- Prefer no exposed ports
Use SSE Server when:
- Need browser access
- Want to test with curl/Postman
- Building web applications
- Need HTTP health checks
- Want monitoring endpoints
📖 Additional Resources
📝 License
MIT
🤝 Support
For issues or questions:
- Check the troubleshooting section
- Review ServiceNow API documentation
- Verify server logs
- Test with curl before using with clients
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。