The Challenge: Getting from Alert to Answer
An alert tells a team that something is wrong. It does not always answer the questions that come next: What happened? Who is affected? Where did the issue originate? What should we do now?
Answering those questions often requires a user to find the relevant test, compare metrics, assess endpoint impact, inspect network paths, and search for related events, outages, or anomalies. The user must then turn that evidence into a clear update for another team or incident channel.
ThousandEyes already provides deep visibility across applications, networks, endpoints, cloud services, SaaS platforms, and third-party dependencies. What customers need is the ability reach the right evidence faster, especially during an active incident or when a non-technical user does not know which dashboard, test, agent, or metric to examine first.
MCP’s Superpower: Building on the Tools You Already Use
MCP gives AI agents a standard way to work with supported ThousandEyes API capabilities. A user asks an AI agent a question in natural language, the agent selects the appropriate MCP tool, and the ThousandEyes MCP server retrieves the requested data through the ThousandEyes API.
For example, a user could ask, “Which tests are failing, and who is affected?” The agent could use MCP tools to find relevant alerts and tests, retrieve performance metrics, assess endpoint impact, and gather supporting path evidence. Users can start with the outcome they need rather than the API endpoint required to retrieve it. This can reduce context switching, shorten troubleshooting time, and accelerate time to resolution.
High-level workflow:
Customer request → AI agent/MCP client → ThousandEyes MCP server → ThousandEyes API
The Bigger Picture
MCP lets teams build on the tools and workflows they already use, without replacing what already works. The following demo brings this flexibility to life, showing how ThousandEyes MCP connects with ServiceNow and Slack to move from investigation to action.
MCP in Action: ThousandEyes MCP + ServiceNow + Slack
This demo shows how an AI agent can use ThousandEyes to first investigate an issue, then create an incident in ServiceNow, and finally share a summary update in Slack:
ThousandEyes investigation → ServiceNow incident → Slack notification
The integration of the ThousandEyes MCP server with ServiceNow and Slack workflow shows just one possibility. MCP’s flexibility opens the door to many more.
Customer Value Delivered
Value Area | How MCP Helps | Value Delivered |
Accessible insights | Users can ask questions in natural language (without requiring knowledge of specific dashboards, metrics, or APIs) while AI agents retrieve relevant ThousandEyes evidence | Lower barrier to entry and easier onboarding, particularly for the majority of MCP users who are new to the technology |
Reduced troubleshooting | Agents can follow repeatable steps to gather alerts, metrics, endpoint data, and path evidence | Faster triage, more complete analysis, and shorter troubleshooting cycles |
Workflow extension and service integration | ThousandEyes evidence can flow into RCA, ServiceNow, Slack, and other operational workflows | Less context switching, richer incident records, and faster team coordination |
Customizable to the customer’s environment | Customers and partners can combine MCP tools and servers with the platforms they already use | Flexible workflows and automation tailored to operational needs |
