MCP Server for xyOps
OfficialClick on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@MCP Server for xyOpsshow me any active alerts and recent job failures"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Connect your AI assistant directly to xyOps. This extension adds a Model Context Protocol (MCP) server at /mcp on your conductor's existing URL, with tools for events, jobs, tickets, alert invocations, monitoring snapshots, and servers.
The extension includes xyCLI as a dependency and runs its MCP HTTP server locally. It works with Docker and manual xyOps installations. You do not need a global CLI installation, a separate container, or an additional published port.
Installation
Requires xyOps 1.1.11 or later.
Install MCP Server from the xyOps Marketplace. Installing the extension enables it automatically after the conductor restarts. No configuration is required for the default setup.
For unpublished development versions, use the local extension development instructions. The package name is @pixlcore/xyext-mcp and its runtime ID is pixlcore-xyext-mcp.
Related MCP server: DoneTick MCP Server
Connect Your AI Assistant
Choose Streamable HTTP in your MCP client and use your conductor's URL with /mcp appended:
https://xyops.example.com/mcpConfigure this request header in your client:
Authorization: Bearer YOUR_XYOPS_API_KEYEach client supplies its own xyOps API key. The extension forwards it to xyCLI, and xyOps validates the key and enforces its permissions and resource restrictions when tools call the API. There is no shared server API key to configure. Choose the permissions you want to grant the assistant when creating its key in xyOps.
To explore the tool list and make calls manually, launch MCP Inspector:
npx @modelcontextprotocol/inspectorSelect Streamable HTTP, enter your conductor's /mcp URL, and configure the Bearer authorization header. See the xyCLI MCP guide for tool descriptions and more client examples.
Configuration
The only setting is port, which defaults to 5524. This is the private port used by the internal service on 127.0.0.1. Clients always connect through the xyOps conductor's /mcp endpoint.
To choose another port, merge this into your xyOps configuration, preserving any existing extension settings:
{
"extensions": {
"pixlcore-xyext-mcp": {
"port": 5525
}
}
}Restart the conductor to apply the change. Choose an unused port between 1 and 65535, different from WebServer.port. In Docker, the equivalent environment option is:
-e XYOPS_extensions__pixlcore-xyext-mcp__port=5525Remove the extension through the Marketplace when you no longer want the MCP endpoint.
Deployment Notes
The child connects back to xyOps at
http://127.0.0.1:PORT. The conductor's HTTP listener must be accessible on loopback.External clients use the conductor's existing HTTPS listener or reverse proxy. Do not publish the internal MCP child port.
Browser requests with an
Originheader must match the origin of xyOps'sbase_app_url. Desktop agents that omit this header can connect normally.Extensions run on the primary conductor. In a multi-conductor cluster with a load balancer, route
/mcpto the current primary.If you use an authentication gateway, allow
/mcpto use Bearer API-key authentication. Browser login redirects cannot authenticate an MCP client.
The extension launches its dependency using the conductor's Node.js executable. Node's module resolution handles both ordinary NPM dependency layouts and development symlinks. Child diagnostics are written to logs/Extension.log with the category column set to pixlcore-xyext-mcp.
The endpoint returns 503 while the child is unavailable. An unexpected child exit is logged and requires a conductor restart; requests are not automatically replayed. During shutdown, the extension removes its URI handler, closes its proxy connections, and sends SIGTERM to the child without waiting for it.
Development Tests
Install the package dependencies and run the tests against a development xyOps checkout:
XYOPS_ROOT=/path/to/xyops npm testWithout XYOPS_ROOT, the tests look for a sibling xyops checkout beside the parent xyplugs directory. They use that checkout's web-server dependencies, the extension's actual xyCLI dependency, and an isolated API fixture. They do not require a running conductor or real API keys.
The tests cover startup and shutdown, custom ports, port conflicts, concurrent client credentials, JSON and SSE forwarding, missing or invalid credentials, rejected origins, and disconnect cleanup.
Privacy
This extension does not collect telemetry or send data to an external service. Tool requests and results travel between your MCP client and xyOps through a local child process. Your AI application may send results to its model provider according to that application's configuration and policies.
License
MIT. See LICENSE.md. xyCLI and its dependencies retain their own licenses.
This server cannot be deployed
Maintenance
Related MCP Connectors
Connect AI assistants to Xitoring monitoring: servers, uptime, incidents, metrics, SSL, and alerts.
Verified, pay-per-use API tools for AI agents through one authenticated connection.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
Exposes FEDLIN's public security scanners as agent-callable tools over Streamable HTTP.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to execute Python functions and system commands via Streamable HTTP, including bash, Python code execution, file operations, and text searching.3,052 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to manage DoneTick chores, subtasks, labels, projects, members, and smart sensor triggers over HTTP SSE.MIT
- FlicenseNot gradedqualityAmaintenanceEnables MCP-compatible AI clients to invoke CLI-driven agent tools over Streamable HTTP, including shell execution, file operations, patching, image viewing, web search, and nested agent tasks, with permission modes and real-time progress streaming.-
- AlicenseNot gradedqualityCmaintenanceEnables AI assistants to search and retrieve support tickets, add notes, close tickets, and access SLA policies through MCP tools, resources, and prompts over Streamable HTTP.MIT