The Model Context Protocol defines a standard method of communication between large language models (LLMs) and external systems, such as a ServiceNow instance. AI applications connect to an external system from an MCP client, such as an AI agent, using an MCP server. The server tells the client which external resources it can access and how to access them. Then, from an MCP client, users can request information from the server and automate functionality using the available tools and data that the server returns. For general information about the Model Context Protocol and MCP servers, see the Model Context Protocol documentation on the Model Context Protocol website.
With MCP Server Console, you can create MCP servers on an instance and configure the tools that they expose or use a preconfigured Quickstart Server. You can create tools from Now Assist skills or use the tools installed with our Quickstart Model Context Protocol Server. Then any LLM or AI agent (such as Moveworks, Microsoft Copilot, or Claude) can call an MCP server from its MCP client using OAuth 2.0 authentication and get a list of available tools. With these tools, users can prompt the server for information from the instance, such as a list of recently closed incidents or case summaries. You can also use the ServiceNow Model Context Protocol Client application as an MCP client.
MCP Server Console benefits
|
Benefit |
Feature |
Users |
|
Integrate with any AI application and MCP client using a standard protocol. |
AI administrator |
|
|
Configure which tools and fields are exposed to MCP clients |
AI administrator |
|
|
Securely access functionality from a ServiceNow instance in any external employee experience. |
AI administrator |
Quickstart Server tools
MCP Server Console includes a preconfigured Quickstart Server to help you get started learning about MCP servers and connecting MCP clients to a ServiceNow instance. The Quickstart Server includes tools for looking up and summarizing incident and case records.Tools define what data an MCP server exposes to clients and are based on Now Assist skills.
|
Name |
Description |
|
Case summarization |
Get a summary of details in a case record generated by Now Assist. |
|
Incident summarization |
Get a summary of details in an incident record generated by Now Assist. |
|
Look up Case Records |
Look up case records to enable summarizing them with the Case Summarization tool. |
|
Look up Incident Records |
Look up incident records to enable summarizing them with Incident summarization tool. |
New
- OAuth Client Registration wizard: register an inbound OAuth client for your MCP server using a guided setup (name and redirect URL) instead of configuring it manually.
- In-console tool testing: list and run any tool directly from the MCP Server Console to validate it before making it available to clients.
- Tool visibility controls: choose which tools are visible in the console and to connected MCP clients.
- Tool annotations: tools now indicate whether they are read-only, destructive, or display a friendly title, so AI clients like Claude can skip confirmation prompts for safe, read-only actions.
- Improved separation between production and sub-production traffic, so testing and development activity no longer competes with production usage.
Changed
- Expanded request tracing, so administrators can follow a single tool call from start to finish.
- Improved throughput and scaling, supporting significantly more concurrent users without performance degradation.
- Enhanced monitoring for rate limits and platform guardrails to help administrators stay ahead of usage issues.
Fixed
- Performance and scaling fixes that address slowdowns under higher concurrent usage.
Removed
- None.
AI Native SKUs of any tier or ProPlus are required to use the MCP Server Console.