Hi ServiceNow Community! 👋
I'm excited to share a tool we've been building that bridges ServiceNow with AI assistants through Anthropic's Model Context Protocol (MCP). This is an open-source project that might be valuable for developers working with ServiceNow automation and AI integration.
What is this?
The ServiceNow MCP Server enables AI assistants (like Claude Desktop) to interact directly with your ServiceNow instances through natural language. Think of it as giving your AI assistant native ServiceNow API access with intelligent context awareness.
Why we built it
As ServiceNow consultants at Happy Technologies, we found ourselves constantly switching between instances (dev, test, prod), manually writing encoded queries, and copy-pasting scripts between our IDE and ServiceNow. We wanted a better way to:
- Work across multiple instances without constant configuration changes
- Use natural language instead of memorizing encoded query syntax
- Develop scripts locally with proper version control
- Automate repetitive ITSM operations
- Discover and work with custom tables automatically
Key Features
🌐 Multi-Instance Support
- Connect to multiple ServiceNow instances simultaneously
- Easy switching: just add "instance": "prod" to any operation
- Centralized configuration in one JSON file
🧠 Intelligent Schema Discovery
- Automatically discovers table structures from your instances
- Works with custom fields and modified tables
- No hardcoded table definitions to maintain
📊 Comprehensive Table Coverage
- 160+ ServiceNow tables including:
- Core ITSM (incident, change_request, problem)
- Service Catalog (sc_cat_item, catalog_ui_policy)
- CMDB & Assets (cmdb_ci_*, alm_asset)
- Platform Development (sys_script, sys_ui_policy, sys_update_set)
- Flow Designer (sys_hub_flow, sys_hub_flow_logic)
- Workflows (wf_workflow, wf_activity)
🎨 Local Script Development
- Sync scripts between local files and ServiceNow
- Watch mode for continuous development
- Full Git integration for version control
- No more copy-paste from Studio!
🗣️ Natural Language Search
- Query using plain English: "all high priority incidents assigned to me"
- Supports 15+ common query patterns
- Translates to encoded queries automatically
⚡ Specialized ITSM Operations
- Convenience tools for common tasks:
- Add comments/work notes
- Assign incidents
- Resolve/close tickets
- Works across Incidents, Changes, and Problems
⚙️ Update Set Management
- Set current update set programmatically
- Move records between update sets
- Clone entire update sets
- Inspect update set contents
Example Use Cases
1. Multi-Instance Incident Management
"Show me all P1 incidents in production, then check if similar issues exist in dev"
2. Update Set Operations
"Switch to my development update set, then move all my changes from Default"
3. Local Script Development
"Watch my ./scripts directory and sync any changes to my dev instance"
4. Natural Language Queries
"Find emergency changes created this week in the prod instance"
5. Batch Operations
"Create 10 test incidents with different priorities for QA testing"
Technical Details
- Built with: Node.js 18+, MCP SDK
- Transports: HTTP (SSE) and Stdio
- Architecture: Metadata-driven, instance-adaptive
- Performance: ~200-500ms per operation, 43+ parallel operations tested
- Authentication: Basic auth with secure credential management
Installation
bash
# Clone and install
git clone https://github.com/Happy-Technologies-LLC/mcp-servicenow-nodejscd mcp-servicenow-nodejsnpm install
# Configure your instances
cp config/servicenow-instances.example.json config/servicenow-instances.json# Edit with your instance details
# Start the server
npm run dev
Full setup instructions available in the README, including Claude Desktop integration.
What's Next?
We're actively developing this and would love community feedback:
- What tables or operations would be most valuable to add?
- What ServiceNow workflows would you automate with AI?
- Any pain points in your current development process this could solve?
Links
🔗 GitHub: https://github.com/Happy-Technologies-LLC/mcp-servicenow-nodejs
📚 Documentation: Comprehensive README and setup guides included
🏢 Company: Part of the Happy Technologies composable service ecosystem
Questions?
Happy to answer any questions about setup, architecture decisions, or use cases. This is an MIT-licensed open-source project, so contributions and feedback are very welcome!
Tags: #Developer #MCP #AI #Automation #API #Integration #OpenSource
