Skip to main content
The Palo Alto Networks Cortex XCOR MCP server supports the tools in the following table. After authenticating, you can query Cortex XCOR and interact with resources from within your code editor.

Get tools

The following tools return individual resources, such as a single dashboard, monitor, or log.

List tools

The following tools list multiple resources, such as a list of all available dashboards, monitors, and log field names.

Query tools

The following tools run a query for a particular data type, such as a query for log data.

Write tools

Write tools modify Cortex XCOR resources. Configure write access before using the tools.

Enable write access

The Cortex XCOR MCP server provides read-only access by default. Enable write access for a connection by adding the X-Chrono-MCP-Read-Only: false header or the ?read_only=false query parameter. Enabling write access exposes available write tools to the MCP client. You can hide specific tools. Select your MCP client for an example:
Include the following option in the claude mcp add command:
The tenant’s allowWritesByDefault: true environment configuration setting enables write access by default. An explicit read-only header or query parameter overrides the tenant setting. Write access doesn’t override the permissions associated with the account or token used to authenticate the connection.

Create and update notebooks

After enabling write access, use the following tools to create or update notebooks: Both tools accept Markdown instead of notebook_json. If you omit collection_slug when creating a notebook, Cortex XCOR uses or creates your user collection. The update_notebook tool replaces the notebook’s complete Markdown content. To avoid overwriting a newer change, pass the updated_at value returned by list_notebooks as expected_updated_at. Both write tools remove a leading level-one heading when it exactly matches the notebook title. The MCP server doesn’t provide a tool to delete notebooks.

Customize available tools

By default, the Cortex XCOR MCP server exposes all available tools to your MCP client. You can use the X-Chrono-MCP-Disable-Tools header to selectively hide specific tools from your client’s tool list. The header accepts a comma-separated list of tool names. Tool names must match the endpoint names listed in the tools table on this page. Whitespace between entries is ignored, and unknown tool names are silently ignored. Select your MCP client for instructions:
To disable specific tools when using Claude Code, include the X-Chrono-MCP-Disable-Tools header when adding the MCP server:
Replace the following:
  • TENANT: The name of your Cortex XCOR tenant.
  • API_TOKEN: The API token generated from your service account. Cortex XCOR recommends storing your API token in a separate file or Kubernetes Secret and calling it using an environment variable, such as $API_TOKEN.
  • TOOLS: A comma-separated list of tool names you want to hide. For example:
To expose all tools, omit the X-Chrono-MCP-Disable-Tools header from the command.