Apollo MCP server

The Apollo Model Context Protocol (MCP) server exposes Apollo CLI functionality as tools for AI agents. You can use the server to enable an AI agent to query environments, manage resources, and search Apollo documentation.

Overview

The Apollo MCP server implements the Model Context Protocol ↗, a standard protocol that enables AI agents to interact with external tools and resources. When you connect an AI agent to the Apollo MCP server, the agent can access approximately 97 tools for managing Apollo resources.

Setup

To connect an AI agent to the Apollo MCP server, use the following connection and authentication details. For instructions on configuring an MCP server, refer to the documentation for your AI agent.

Connection details

• Transport type: http (HTTP with Server-Sent Events)
• Server URL: https://<apollo-hub-url>/apollo-mcp/mcp
  • Replace <apollo-hub-url> with the URL of your Apollo Hub.

Authentication

The Apollo MCP server uses an Apollo Bearer Token for authentication. To create a token, follow the instructions to create a Bearer Token.

Pass the Bearer Token in the Authorization header when you connect to the MCP server:

Copied!
1
Authorization: Bearer <your-apollo-token>

Example configurations

The following examples provide the connection details in common configuration formats. Adapt an example to the format that your AI agent requires.

The following example uses JSON:

Copied!
1
2
3
4
5
6
7
{
  "type": "http",
  "url": "https://apollo.palantircloud.com/apollo-mcp/mcp",
  "headers": {
    "Authorization": "Bearer <your-token>"
  }
}

The following example uses YAML:

Copied!
1
2
3
4
type: http
url: https://apollo.palantircloud.com/apollo-mcp/mcp
headers:
  Authorization: "Bearer <your-token>"

Verification

After you configure the MCP server, verify the connection in your terminal:

1. Ask the agent to list the available Apollo tools. The agent should return approximately 97 tools.
2. Ask the agent to perform a simple query, such as "Search the Apollo documentation for environments."
3. Confirm that the agent can call apollo_docs_search successfully.

Discovery tools

The Apollo MCP server provides tools that help AI agents discover available Apollo resources and structure queries.

Search the schema

The Apollo GraphQL API is the primary interface for querying Apollo resources. The MCP server includes a schema search tool that enables AI agents to find queries, mutations, types, fields, and enumeration values in the GraphQL schema.

Agents can use this tool to:

• Find the query or mutation for a task.
• Discover the available fields on a type.
• Understand the input parameters for an operation.
• Find the enumeration values for a field.

Apollo documentation search

The MCP server includes embedded Apollo documentation that AI agents can search and read. The following documentation tools are available:

• apollo_docs_search: Searches the Apollo documentation.
• apollo_docs_read: Reads a specific section of the Apollo documentation.

Agents can use these tools to:

• Understand Apollo concepts, such as Environments, Modules, and Release Channels.
• Learn about workflows, such as deploying Modules and managing change requests.
• Find best practices, such as when to use Modules instead of Entities.
• Troubleshoot issues and understand error messages.