Answers with context
Let an assistant look up an order, check its status, or find a record using tools connected to your team’s API.
“Find order ORD-1042”
CONNECT YOUR APIs TO MCP
Give your AI tools access to the services you already use. Turn an API specification into an MCP server, choose what to expose, and connect your preferred client.
Bring your API. Keep your existing backend.
- OpenAPIOpenAPI 3.x & Swagger 2.0
- You control accessScoped access keys
REST API MCP AI CLIENT
EXPLORE A CONNECTION. BEFORE YOU BUILD ONE.01 — 03
EXAMPLE WORKSPACE
ServersCommerce API
Commerce API
api.acme.exampleTool / endpointMethodEnabled
Tool input
Look up an order by its ID.
{
"order_id": "ORD-1042"
}What’s the status of order ORD-1042?
get_orderOrder ORD-1042 has been delivered.Illustrative example. Select a tool to inspect its input and sample response.
CONNECT TO YOUR PREFERRED MCP CLIENT
ClaudeCursorVS Coden8nYour own client
A useful connection.
In three steps.
Your services already know how to do the work. Make their capabilities available through a standard your AI tools understand.
- 01
Bring your API
Import an OpenAPI or Swagger file, paste a specification URL, or define your endpoints manually.
- 02
Choose the tools
Review generated tools, refine their descriptions, and enable the endpoints your client should use.
- 03
Test, then connect
Try a tool in the console, inspect the response, and add the server URL to your MCP client.
An adapter for your AI tools.
MCP (Model Context Protocol) is a standard way for AI assistants to use external tools. Wrap-to-MCP connects that standard to your REST APIs, so an assistant can request useful data or carry out a task.
Your APIServices you already run
Wrap-to-MCPTools you choose to expose
Your AI clientClaude, Cursor, and more
Your API handles the work. Your client decides when to call a tool.
Technical underneath.
Useful to everyone.
Give developers a manageable integration and give your team a simpler way to use it.
Fewer manual handoffs
Make routine API actions available as named tools. Your client can use them in the workflow you’re already in.
“Create a support ticket”
A shared place to work
Organize API servers in workspaces and work with your team to review the tools available to each client.
One workspace. Your team’s tools.
A standard endpoint.
A familiar setup.
Connect over Streamable HTTP. Use a server endpoint for one API, or a workspace endpoint to bring multiple servers together.
- Structured tool inputs
Review the generated JSON schemas in your console.
- Access at the right scope
Configure workspace, server, or single-tool API keys.
- A way to test every tool
Check arguments and inspect actual API responses.
.cursor/mcp.json
{
"mcpServers": {
"commerce": {
"url": "https://your-host/mcp/acme/commerce",
"headers": {
"X-API-Key": "wtm_your_key_here"
}
}
}
}Replace the host, workspace, server prefix, and key with your own. Client support and setup vary; check the connection guide.
A plan for your
next connection.
From an individual API project to shared team workflows. Choose the setup that fits how you work.
FOR INDIVIDUAL PROJECTS
Starter
$19/ month
A starting point for connecting your API to an MCP client.
- OpenAPI, Swagger, or manual endpoints
- Tool editing and browser testing
- Streamable HTTP connections
FOR SHARED INTEGRATIONS
Team
$59/ month
Organize your team’s APIs and manage access in one place.
- Shared workspaces
- Multiple API servers in a workspace
- Workspace, server, and tool API keys
FOR YOUR DEPLOYMENT NEEDS
Enterprise
Custom
For teams with their own infrastructure and access requirements.
- Self-hosting setup options
- Configurable API access
- Workspace and membership controls
Prefer to self-host?
Run the console, backend, and database on your own infrastructure. Hosting, database operations, and any API or AI-provider charges are paid separately by you.