UnifAPI MCP Server

MCP serverAI & models

Give your AI live access to public data and a set of skills built for agents. Once added, it can pull current information from public-data APIs whenever a task calls for it and use the included skills to get more done. Everything is hosted for you, so there is nothing to install or run yourself.

Available today. Use it from your connected AI after setup.

No other account needed.

After adding it, ask your AI something that depends on current public data and see what it brings back. Then try one of the included skills to explore what else it can handle.

Then ask your AI: use UnifAPI MCP Server

What your AI can do with it

Fetch live data from public APIs
Answer questions using current public information
Use included skills made for AI agents
Pull fresh public data whenever a task needs it
Start right away since it is hosted and ready to use

From the project's README

As published by unifapi-agent/unifapi-mcp-server in README.md.

UnifAPI is a hosted MCP server for AI agents that need live public-data APIs. It connects Codex, Claude Code, Cursor, and other Model Context Protocol clients to task-focused Skills for creator research, social listening, competitor monitoring, KOL pricing, and public web data workflows.

MCP server

Repository: unifapi-agent/unifapi-mcp-server
Registry name: com.unifapi/mcp
Transport: Streamable HTTP
Server URL: https://mcp.unifapi.com
MCP setup: unifapi.com/mcp
Website: unifapi.com
API catalog: unifapi.com/apis
Skills catalog: unifapi.com/skills
Documentation: docs.unifapi.com/mcp

Quick install

Codex:

codex mcp add unifapi --url https://mcp.unifapi.com

Claude Code:

claude mcp add --transport http unifapi https://mcp.unifapi.com

Cursor:

{
  "mcpServers": {
    "unifapi": {
      "url": "https://mcp.unifapi.com"
    }
  }
}

What agents can do

UnifAPI exposes three MCP tools:

list_operations - search the public-data operation catalog.
get_operation - inspect parameters, schemas, and call examples.
call_api - call a concrete operation and receive data plus billing metadata.

Common workflows include:

X (Twitter) search, creator lookup, and KOL pricing research.
TikTok, YouTube, Instagram, Reddit, and LinkedIn-style public-data analysis.
Influencer discovery and creator shortlist generation.
Competitor monitoring from public posts, comments, profiles, and trends.
Market research and lead enrichment for AI agents.

Authentication

OAuth is the recommended authentication path. Compatible MCP clients discover the protected resource metadata from:

https://mcp.unifapi.com/.well-known/oauth-protected-resource

Clients that cannot complete OAuth can use a UnifAPI API key as a bearer token. See the MCP docs for current setup details.

Keywords

Model Context Protocol, MCP server, remote MCP server, hosted MCP server, AI agent tools, public data API, social media API, Twitter API, X API, TikTok API, YouTube API, Instagram API, Reddit API, LinkedIn API, creator research, influencer discovery, KOL pricing, competitive intelligence, social listening, UnifAPI.

Signals

GitHub stars: 3
Forks: 1
Last commit: May 2026

Advanced

Delivery: mcp MCP server → your ahel gateway (mcp.ahel.ai) → every connected AI client.
Catalog kind: mcp-server
Gateway key: com-unifapi-mcp
Source: github.com/unifapi-agent/unifapi-mcp-server
Hosted endpoint: https://mcp.unifapi.com

github.com/unifapi-agent/unifapi-mcp-server

More in AI & models

atom-mcp-server
MCP server · a7om-ai ottasia
MCP server · aiweather-anurag unphurl
MCP server · 123ergo free-blueprints
MCP server · agent-blueprint context7
MCP server · upstash tldraw
MCP server · tldraw