# AI Hub

> Use this tool when you need to access and manage multiple AI services from a single interface, solving vendor lock-in and model management problems for developers and organizations. It takes in configuration files and input queries, outputting model information and chat responses, and supports various transport options and deployment configurations. Ideal for use cases requiring flexibility and centralized management across AI providers like OpenAI, Google, and AWS.

Canonical page: https://skillsregistry.net/skills/feiskyer-ai-hub  
JSON: https://api.skillsregistry.net/v1/skills/feiskyer-ai-hub

## Description

This MCP AI Hub server by Pengfei Ni provides unified access to 100+ AI providers through LiteLLM integration, enabling seamless switching between OpenAI, Anthropic, Google, Azure, AWS Bedrock, and other AI services through a single configuration file. Built with Python using FastMCP and featuring comprehensive model management with YAML-based configuration, environment variable support, and robust error handling, it offers three core tools for chatting with models, listing available models, and retrieving model information with support for both string and OpenAI message format inputs. The implementation includes extensive testing coverage, multiple transport options (stdio, SSE, HTTP), and flexible deployment configurations, making it ideal for developers building AI applications that need provider flexibility, organizations wanting to avoid vendor lock-in, and teams requiring centralized AI model management across different services without code changes.

## Trust

- **Trust score (0–1):** 0.22
- **Verification tier:** scanned
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** cloud-infra
- **Updated:** 2026-09-28

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/feiskyer-ai-hub)
- **Repository:** <https://github.com/feiskyer/mcp-ai-hub>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "feiskyer-ai-hub"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/feiskyer-ai-hub` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/feiskyer-ai-hub/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt