Installation

Connect Reqistry to your AI coding agent with MCP, then ask it to find, compare, inspect, and submit APIs while you work.

Overview

Reqistry is an API intelligence platform for developers and AI agents. It catalogs 1,500+ APIs across 50+ categories and adds practical signals such as reliability, authentication model, pricing transparency, CORS support, documentation quality, and agent-readiness.

The web and mobile apps are for people who want to browse the catalog directly. The MCP server is for coding agents such as Claude Code, Codex, OpenCode, Cursor, and Windsurf.

Endpoint: https://mcp.reqistry.dev/mcp. No Reqistry sign-up or API key is required.

Installation

Pick the agent you use, add the Reqistry MCP server, then restart or refresh the agent so it reloads MCP tools.

Claude Code

Recommended for most Claude Code users. The default local scope keeps the server private to your current project.

claude mcp add --transport http reqistry https://mcp.reqistry.dev/mcp

To share the configuration with a team, add --scope project before reqistry.

OpenAI Codex

Add this to ~/.codex/config.toml for all projects, or .codex/config.toml for a trusted project.

[mcp_servers.reqistry]
url = "https://mcp.reqistry.dev/mcp"

In Codex CLI, run codex mcp list to verify it was loaded, or use /mcp inside the Codex TUI.

OpenCode

OpenCode v2 keeps MCP servers under mcp.servers.

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "servers": {
      "reqistry": {
        "type": "remote",
        "url": "https://mcp.reqistry.dev/mcp",
        "oauth": false
      }
    }
  }
}

Cursor

Create .cursor/mcp.json for one project or ~/.cursor/mcp.json for global use.

{
  "mcpServers": {
    "reqistry": {
      "type": "http",
      "url": "https://mcp.reqistry.dev/mcp"
    }
  }
}

Windsurf

Open Windsurf settings, choose Manage MCPs, then edit the raw MCP config. On macOS and Linux the file is usually ~/.codeium/windsurf/mcp_config.json.

{
  "mcpServers": {
    "reqistry": {
      "serverUrl": "https://mcp.reqistry.dev/mcp"
    }
  }
}

Any MCP-compatible client

Use a remote Streamable HTTP MCP server with this endpoint:

https://mcp.reqistry.dev/mcp

What You Can Do

Reqistry MCP gives your agent structured API intelligence it can use while planning or implementing work. You do not need to learn commands; ask in plain language.

Discover APIs by requirement

Ask for APIs that match a concrete product need, such as free weather data, financial market data, translation, geocoding, screenshots, or email delivery.

Example: Find reliable free weather APIs that do not require OAuth.

Evaluate integration risk

Review availability, authentication model, pricing category, documentation quality, browser compatibility, and other signals before committing engineering time.

Example: Compare OpenWeatherMap alternatives and tell me which one is safest for a browser app.

Use API intelligence inside your workflow

Let your coding agent research options while it is planning code, writing integration tasks, or reviewing whether an API is still usable.

Example: For this project, recommend five APIs for sending transactional email and explain the trade-offs.

Inspect individual APIs

Pull a structured profile for a known API, including category, source links, health signals, and practical usage notes where available.

Example: Show me the Reqistry profile for OpenWeatherMap and check its health.

Submit missing APIs

Send new API candidates into the Reqistry review flow so the catalog improves over time instead of staying static.

Example: Submit this API to Reqistry: https://example.com/api.

Browse the catalog

Explore categories and API counts when you are not sure what kind of service exists for a problem yet.

Example: List the API categories and suggest categories relevant to travel planning.

FAQ

What is MCP?

MCP, or Model Context Protocol, is a standard for connecting AI clients to external tools and data. Reqistry exposes an MCP endpoint so compatible coding agents can search and inspect the API catalog during normal conversation.

Do I need an account or API key?

No. Reqistry MCP is public and does not require a Reqistry account, API key, or credit card.

Which agents are supported?

Reqistry works with clients that support remote MCP over Streamable HTTP. The examples on this page cover Claude Code, OpenAI Codex, OpenCode, Cursor, and Windsurf.

Should I use the web app, mobile app, or MCP?

Use the app when you want to browse and compare APIs yourself. Use MCP when you want your coding agent to research APIs while it plans, writes, or reviews code.

Why do the config examples look different per agent?

Each agent stores MCP servers in its own format. The endpoint is the same; only the client configuration changes.

How do I know it connected?

Restart or refresh your agent after saving the config, then open its MCP tools view if it has one. Ask for Reqistry categories or a specific API profile to confirm the tools are available.

How often is the data updated?

Reqistry continuously maintains the catalog with health checks and review workflows. Results should be treated as decision support, not as a replacement for reading a provider's official API terms before production use.

Can I submit my own API?

Yes. Ask your agent to submit the API URL and basic details. Submissions enter a review flow before publication.