MCP Call

Tip

This document describes the detailed usage of a specific policy. If you are using the Apinizer policy structure for the first time or want to learn how policies work, read What is Policy? first.

Overview

What is its Purpose?

• Calls the MCP tool selected by the administrator on every matching request.
• Does not leave tool selection to the LLM.
• Resolves the argument template, validates constraints, and then invokes the tool.

Working Principle

1. The policy connects to the MCP server endpoint.
2. Capability discovery returns the tool catalog.
3. targetToolName is selected as the fixed target.
4. argumentsTemplate is resolved into JSON arguments.
5. Argument constraints are validated and the tool is called.

Features and Capabilities

Basic Features

• Fixed tool: targetToolName is required on the Fixed variant.
• Argument template: Must produce valid JSON after variables are resolved.
• Constraints: REQUIRED, REGEX, ENUM, and MAX_LENGTH can be applied.
• Fail-closed: An empty discovery result or allow-list does not make every tool available.

Advanced Features

• Connection: Endpoint and transport must match the MCP server. Authentication, mTLS, and private egress are configured for the target environment.
• Cache: Capability Cache TTL is how often the tool catalog is refreshed.
• Drift: FAIL_CLOSED stops the call when the catalog changes.
• Error: connectionConfigMcpRef or targetToolName not configured means the server or target tool is missing from the runtime configuration.

Usage Scenarios

Scenario	Situation	Solution (Policy Application)	Expected Behavior / Result
Fixed customer search	The same tool must run on every support-assistant request	search_customer is selected; the email from the header is sent through the JSON argument template	The result is written to a gateway variable

Configuring Policy Parameters

MCP Server and Discovery

• Endpoint and transport must match the MCP server.
• Configure authentication, mTLS, and private egress for the target environment.
• After discovery, the tool count must be greater than zero.

Target Tool and Arguments Template

• targetToolName is required on the Fixed variant.
• The template must produce valid JSON after variables are resolved.
• REQUIRED, REGEX, ENUM, and MAX_LENGTH constraints can be applied.
• Confirm that the target tool selection is kept on the edit screen.

Cache and Drift

• Capability Cache TTL is how often the tool catalog is refreshed.
• FAIL_CLOSED stops the call when the catalog changes.

Fail-Closed Behavior

An empty discovery result or allow-list does not make every tool available.

Checklist

• Are endpoint and transport correct?
• Are authentication and mTLS correct?
• Did tool discovery succeed?
• Was the target tool selected and saved?
• Does the arguments template produce valid JSON?
• Is the selection kept on the edit screen?
• Was the API Proxy redeployed?

Deleting the Policy

For deletion steps and what to do while the policy is in use, see Policy Management.

Exporting and Importing the Policy

For export and import steps, see Export/Import.

Connecting the Policy to an API

For how to attach this policy to APIs, see Policy Management.