DISCOVER THE FUTURE OF AI AGENTSarrow_forward

Apollo MCP Server

calendar_todayAdded Feb 26, 2026
categoryAgent & Tooling
codeOpen Source
Workflow AutomationRustModel Context ProtocolAI AgentsCLIAgent & ToolingModel & Inference FrameworkDeveloper Tools & CodingProtocol, API & Integration

An official Rust-based MCP server by Apollo that converts GraphQL APIs into standardized AI-callable tools, supporting multi-client integration like Claude and Cursor for intelligent LLM and graph data orchestration.

Overview#

Apollo MCP Server is a high-performance MCP server developed by Apollo GraphQL using Rust (98.6%). It serves as a middleware adapter that transforms GraphQL API operations (queries, mutations) into tools compliant with the Model Context Protocol (MCP), enabling AI models to access and orchestrate APIs in a standardized manner.

Core Value: Solves the challenge of LLMs struggling to directly understand and use complex GraphQL APIs, eliminating the need for custom integrations for each AI platform.

Core Features#

  • GraphQL to MCP Conversion: Automatically maps GraphQL operations to MCP Tool definitions
  • Multiple Operation Sources: Operation Collections, Uplink manifests, local files, direct definitions
  • Multi-Client Compatibility: Claude Desktop, Claude Code, Cursor, Goose, Cline, Windsurf, OpenCode
  • Built-in Introspection Tools: execute, introspect, search, validate with configurable hints
  • Security Mechanisms: Host header validation against DNS rebinding attacks, multiple authentication methods
  • Observability: Configurable metrics export intervals, OpenTelemetry support
  • Protocol Version Negotiation: Dynamic enablement of outputSchema and structuredContent
  • Custom Configuration: Customizable server metadata and tool descriptions

Quick Start#

Prerequisites: Rover CLI v0.37+, Node.js v18+, MCP-compatible client

Initialize Project:

rover init --mcp

Run Server:

rover dev --supergraph-config supergraph.yaml --mcp .apollo/mcp.local.yaml

Verify Server:

npx @modelcontextprotocol/inspector http://127.0.0.1:8000/mcp --transport http

Docker Deployment:

docker build -f mcp.Dockerfile -t my-mcp-server .
docker run -p 4000:4000 -p 8000:8000 \
  -e APOLLO_KEY=$APOLLO_KEY \
  -e APOLLO_GRAPH_REF=$APOLLO_GRAPH_REF \
  -e MCP_ENABLE=1 \
  my-mcp-server

Configuration Examples#

Server Metadata:

server_info:
  name: "Acme Corp GraphQL Server"
  version: "2.0.0"
  title: "Acme MCP Server"
  website_url: "https://acme.com/mcp-docs"

Custom Tool Descriptions:

overrides:
  descriptions:
    GetAlerts: "Get active weather alerts for a US state"

Security Configuration:

transport:
  type: streamable_http
  host_validation:
    enabled: true
    allowed_hosts:
      - mcp.example.com

Architecture Highlights#

  • Core Language: Rust (98.6%), high performance and memory safety
  • Dependency: rmcp (v0.14+) for underlying MCP protocol implementation
  • Apollo Router Integration: Runs as extension or standalone process
  • Transport Layer: Streamable HTTP transport with SSE Resumability

Use Cases#

  • Enterprise AI assistants integrating internal GraphQL microservices
  • AI-powered intelligent orchestration of complex multi-API workflows
  • Building unified AI tool gateways supporting multiple clients
  • Rapid prototyping and local development testing

Related Projects

View All arrow_forward

STAY UPDATED

Get the latest AI tools and trends delivered straight to your inbox. No spam, just intelligence.

rocket_launch