# MCP Aggregator Servers: 10 Production Examples from awesome-mcp-servers > Discover 10 real-world MCP Aggregator server examples from awesome-mcp-servers. Simplify AI agent interactions by unifying multiple tools into a single endpoint. - Repository: [Frank Fiegel/awesome-mcp-servers](https://github.com/punkpeye/awesome-mcp-servers) - Tags: getting-started - Published: 2026-09-02 --- **MCP Aggregator servers consolidate multiple tools, services, or downstream MCP servers into a single unified endpoint, allowing AI agents to discover and invoke hundreds of capabilities through one interface without managing individual server connections.** The `punkpeye/awesome-mcp-servers` repository maintains a curated registry of Model Context Protocol implementations, including a dedicated **Aggregators** section in [[`README.md`](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md)](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md#aggregators) that catalogs servers designed to unify disparate tools. These MCP Aggregator servers eliminate integration complexity by exposing a consolidated `tools/list` interface and handling routing, billing, and schema normalization automatically. ## What Is an MCP Aggregator Server? An MCP Aggregator server acts as a meta-layer that exposes many tools or other MCP servers through a single HTTP or stdio endpoint. According to the source analysis of the awesome-mcp-servers registry, these aggregators share four key architectural traits: - **Unified Interface:** A single endpoint internally forwards calls to a collection of downstream tools or servers, abstracting away multiple connection points. - **Dynamic Discovery:** Aggregators expose a `tools/list` meta-tool that lets agents query available capabilities at runtime without hardcoding endpoints. - **Pay-Per-Call Billing:** Many implement the x402 micropayment model, charging USDC per invocation rather than requiring API keys or subscriptions. - **Cross-Platform Support:** Implementations exist in TypeScript (📇), Python (🐍), Go (🏎️), and Rust (🦀), enabling integration into diverse environments. ## 10 MCP Aggregator Server Examples The following aggregators are cataloged in the [Aggregators section](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md#aggregators) of the awesome-mcp-servers repository: 1. **Correctover/mcp-server** – As documented at line 136 of the registry's [`README.md`](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md), this TypeScript aggregator provides contract validation and self-healing failover for LLM APIs. It features 6-dimensional verification, 87 self-healing rules, and Bring-Your-Own-Key (BYOK) support to 9 providers with sub-millisecond latency, supporting both cloud and local deployments. 2. **daedalusdevelopmentgroup/ddg-agent-payable-services** – Listed at line 137, this Python-based pay-per-call gateway aggregates 90+ agent tools. It uses USDC x402 billing and installs via `pip install ddg-agent-services-mcp`. 3. **1mcp/agent** – Referenced at line 140, this TypeScript implementation aggregates multiple downstream MCP servers into a unified endpoint. It supports cross-platform operation on macOS, Windows, and Linux, functioning in both cloud and local modes. 4. **2s-io/sdk** – Documented at line 141, this TypeScript aggregator provides 180+ tools spanning weather, geocoding, legal, and finance via a single MCP API. It uses x402 micropayments, requires no API keys, and launches with `npx -y @2sio/mcp`. 5. **mcpqueen/mcpqueen** – Found at line 144, this TypeScript aggregator maintains a graded registry that live-probes every official MCP server. It exposes a searchable endpoint via `mcpqueen.com` with quality scoring capabilities. 6. **Work90210/APIFold** – Listed at line 146, this TypeScript tool converts any REST API into a hosted MCP server, currently exposing 18 public APIs with zero-setup deployment and no API key requirements. 7. **agentbodegastore/agentbodega** – Documented at line 148, this TypeScript aggregator maintains a live catalog of 65 x402-payable HTTP tools. It installs via `npx -y @agentbodega/mcp` and focuses on monetized tool access. 8. **glena-AI/awesome-mcp-tools-mcp** – Referenced at line 165, this MIT-licensed TypeScript tool provides a CLI and stdio bridge to the awesome-mcp.tools catalog containing 2,000+ MCP servers. It supports commands like `npx awesome-mcp search ` with zero dependencies. 9. **Markgatcha/universal-mcp-toolkit** – Found at line 199, this TypeScript toolkit aggregates multiple MCP servers through a unified configuration system. It provides ready-made templates and cross-repository prompt workflows for both cloud and local use. 10. **MCPJungle (duaraghav8)** – Listed at line 172, this Go-based aggregator provides a self-hosted registry designed for enterprise AI agents. It supports private, on-premise deployment for internal agent fleets. ## How to Deploy MCP Aggregator Servers Each aggregator implements the Model Context Protocol differently, but most provide CLI-based initialization. Below are deployment patterns for three representative implementations from the awesome-mcp-servers registry. ### TypeScript Aggregator: Correctover The Correctover aggregator launches a local HTTP endpoint that proxies to multiple LLM providers. After installation, it exposes the standard MCP interface on port 3000. ```bash npx -y correctover-mcp-server ``` Once running, query the available tools via the `/mcp` endpoint: ```bash curl -X POST http://localhost:3000/mcp \ -d '{"method":"tools/list","params":{}}' | jq ``` ### Python Aggregator: Daedalus Development Group This Python-based aggregator implements x402 billing for 90+ tools. Install the package and start the server on port 8080. ```bash pip install ddg-agent-services-mcp python -m ddg_agent_services_mcp ``` Invoke a specific tool by posting to the MCP endpoint: ```bash curl -X POST http://localhost:8080/mcp \ -d '{"method":"price/get","params":{"symbol":"BTC"}}' | jq ``` ### CLI Aggregator: awesome-mcp-tools The `glena-AI/awesome-mcp-tools-mcp` package provides a searchable bridge to 2,000+ servers. Search for capabilities, then launch a meta-server. Search for weather-related MCP servers: ```bash npx awesome-mcp search weather ``` Start the aggregator that merges discovered results: ```bash npx awesome-mcp serve ``` With the server running on port 4000, call any discovered tool directly: ```bash curl -X POST http://localhost:4000/mcp \ -d '{"method":"weather/get","params":{"location":"Paris"}}' | jq ``` ## Summary - **MCP Aggregator servers** expose hundreds of downstream tools through a single unified endpoint, abstracting connection management for AI agents. - They typically implement **dynamic discovery** via `tools/list` methods and support both cloud and local deployment modes. - **x402/USDC pay-per-call billing** is common among aggregators like Daedalus and 2s-io, eliminating API key management. - The **awesome-mcp-servers** repository catalogs 10+ production aggregators across TypeScript, Python, and Go, ranging from contract-validated LLM gateways to enterprise self-hosted registries. ## Frequently Asked Questions ### What is the difference between a standard MCP server and an MCP Aggregator? A standard MCP server typically exposes a single tool or API resource, while an MCP Aggregator unifies multiple servers or tools under one endpoint. Aggregators handle internal routing, schema normalization, and often implement discovery protocols like `tools/list` to let agents browse available capabilities at runtime. ### How does pay-per-call billing work in MCP Aggregators? Aggregators implementing the x402 protocol charge micropayments in USDC for each tool invocation rather than requiring monthly subscriptions or API keys. For example, the Daedalus Development Group aggregator bills per call for its 90+ tools, while AgentBodega exposes 65 HTTP tools under the same payment model. ### Can MCP Aggregator servers run entirely on-premise? Yes. Several aggregators support private, self-hosted deployment. MCPJungle (written in Go) is specifically designed for enterprise on-premise registries, while Correctover and 1mcp/agent support local execution on macOS, Windows, and Linux without cloud dependencies. ### Where can I find the complete list of MCP Aggregator servers? The canonical list resides in the **Aggregators** section of the [[`README.md`](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md)](https://github.com/punkpeye/awesome-mcp-servers/blob/main/README.md#aggregators) file within the `punkpeye/awesome-mcp-servers` repository. This section catalogs each aggregator's language, deployment options, and unique features such as billing models and verification capabilities.