Documentation MCP Server - Connexease Gateway

The Connexease Gateway documentation is available as a hosted Model Context Protocol (MCP) server. Once connected, AI assistants such as Claude, Cursor, VS Code Copilot, Windsurf, Codex, and Gemini CLI can search and read these docs directly, so their answers about the Messaging API, Public API, Embedded Signup SDK, and the Gateway Dashboard are grounded in the current, official documentation instead of the model’s training data.

https://docs.gateway.connexease.com/mcp

The documentation MCP server is public and read-only. It requires no account, sign-in, or API key, and it cannot read or change anything in your Gateway account. To manage applications, templates, webhooks, or sandbox numbers from an AI assistant, use the Gateway Customer MCP plugin instead.

What it provides

Tool	What it does
`search_connexease_gateway`	Semantic search across all documentation pages. Returns matching sections with their titles, content, and direct page links. Best for conceptual questions such as “how do I authenticate?”
`query_docs_filesystem_connexease_gateway`	Read-only, shell-like access (`rg`, `cat`, `head`, `tree`, `jq`, …) to the docs pages and the OpenAPI specs (`/openapi/openapi.json`, `/openapi/public-openapi.json`). Best for exact keyword matches, reading a full page, or inspecting endpoint schemas.
`submit_feedback`	Reports an incorrect, outdated, or unclear documentation page to the docs team.

Tool names may appear with a slightly different prefix depending on your client; the server and behavior are the same.

Connect your AI tool

The server uses the Streamable HTTP transport. Pick your client below.

Open Settings → Connectors.
Click Add custom connector.
Enter a name, for example Connexease Gateway Docs, and the URL https://docs.gateway.connexease.com/mcp.
Click Add. No authentication is required.
In a chat, open the tools menu and make sure the connector is enabled.

On Team and Enterprise plans, an organization owner may need to add the connector first.

Run this in your terminal:

claude mcp add --transport http connexease-gateway-docs https://docs.gateway.connexease.com/mcp

Add --scope user to make it available in every project, or --scope project to share it with your team through .mcp.json. Verify with:

claude mcp list

Inside a Claude session, /mcp shows the connection status.

Add the server to ~/.cursor/mcp.json (global) or .cursor/mcp.json (project):

{
  "mcpServers": {
    "connexease-gateway-docs": {
      "url": "https://docs.gateway.connexease.com/mcp"
    }
  }
}

Then open Cursor Settings → MCP and confirm the server is enabled.

Add the server to .vscode/mcp.json in your workspace:

{
  "servers": {
    "connexease-gateway-docs": {
      "type": "http",
      "url": "https://docs.gateway.connexease.com/mcp"
    }
  }
}

Open Copilot Chat in Agent mode and enable the server’s tools from the tools picker.

Add the server to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "connexease-gateway-docs": {
      "serverUrl": "https://docs.gateway.connexease.com/mcp"
    }
  }
}

Refresh the MCP servers list in Cascade.

Add the server to ~/.codex/config.toml:

[mcp_servers.connexease-gateway-docs]
url = "https://docs.gateway.connexease.com/mcp"

Restart Codex and run /mcp to confirm the tools are listed.

Add the server to ~/.gemini/settings.json:

{
  "mcpServers": {
    "connexease-gateway-docs": {
      "httpUrl": "https://docs.gateway.connexease.com/mcp"
    }
  }
}

Restart Gemini CLI and run /mcp to see the server and its tools.

If your client only supports local (stdio) servers, bridge the remote server with mcp-remote (requires Node.js 18+):

{
  "mcpServers": {
    "connexease-gateway-docs": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://docs.gateway.connexease.com/mcp"]
    }
  }
}

Usage examples

After connecting, ask in natural language. Mentioning “Connexease Gateway docs” in your prompt helps the assistant pick the right tool.

Example prompts

According to the Connexease Gateway docs, how do I send a template message with body variables? Give me a cURL example and the page link.

What is the difference between the Messaging API and the Public API in Connexease Gateway?

How can I test a request in the Sandbox without sending a real WhatsApp message?

What does error code BILL_001 mean and how do I avoid it?

List every Public API endpoint for templates from the OpenAPI spec, with method and path.

Which webhook event do I receive when a user taps a quick reply button? Show the payload.

What are the rate limits per API key, and which response headers should I watch?

Write a Node.js Express handler that receives Gateway webhooks and verifies them as described in the Authentication page.

Example conversation

The following is an illustrative flow. Responses are summarized; the assistant typically also includes code samples and links.

❯ Using the Connexease Gateway docs, how do I send a WhatsApp template message with two body variables?

⏺ [search_connexease_gateway: "send template message body variables"]

  Send a POST request to /v1/wa/message with type "template". Put the approved
  template name and language code in "template", and the variable values in a
  "body" component:

  curl -X POST https://api.gateway.connexease.com/v1/wa/message \
    -H "Authorization: Bearer <API_KEY>" \
    -H "Content-Type: application/json" \
    -d '{
      "messaging_product": "whatsapp",
      "to": "905321234567",
      "type": "template",
      "template": {
        "name": "shipping_update",
        "language": { "code": "en_US" },
        "components": [
          { "type": "body", "parameters": [
            { "type": "text", "text": "ORD-20251203" },
            { "type": "text", "text": "In Transit" }
          ]}
        ]
      }
    }'

  Source: https://docs.gateway.connexease.com/api-reference/send-message

❯ From the Public API OpenAPI spec, which endpoints exist under /wa/{app_id}/templates?

⏺ [query_docs_filesystem_connexease_gateway:
   cat /openapi/public-openapi.json | jq '.paths | keys']

  The template endpoints are list, create, get detail, update, delete, and metrics.
  See the Template section of the Public API reference for request and response details.

Tips

Ask for the source link. Adding “include the documentation link” to your prompt makes it easy to verify the answer.
Be specific about the API. Say “Messaging API” (sending and receiving messages) or “Public API” (templates, analytics, business profile, webhooks, API keys) to get the right reference.
Use the OpenAPI specs for exact schemas. Ask the assistant to read /openapi/openapi.json or /openapi/public-openapi.json when you need exact field names, enums, or response shapes.
Never paste secrets. The server does not need your API key; keep sk_ and pk_ keys out of chat prompts.

Troubleshooting

Problem	What to try
The tools do not appear	Restart the client after adding the server, and make sure the URL ends with `/mcp`.
The client asks for authentication	No authentication is needed. Remove any headers or OAuth settings you added and reconnect.
Connection fails with a transport error	Your client may not support Streamable HTTP; use the `mcp-remote` bridge shown in Other clients.
The assistant answers without using the docs	Mention “Connexease Gateway docs” in your prompt, or explicitly ask it to search the documentation first.
A page looks wrong or outdated	Ask the assistant to send feedback with `submit_feedback`, or contact Connexease support.

​What it provides

​Connect your AI tool

​Usage examples

​Example prompts

​Example conversation

​Tips

​Troubleshooting

What it provides

Connect your AI tool

Usage examples

Example prompts

Example conversation

Tips

Troubleshooting