Connecting Dwizi Tools Anywhere

Dwizi tools are callable capabilities exposed through standard interfaces. This allows them to work across AI clients, automation platforms, and APIs.

Once connected, tools can be discovered, invoked, and governed consistently.

AI Clients (MCP)

These clients connect using Model Context Protocol (MCP). After setup, all tools in your Dwizi workspace become available automatically.

Cursor

1. Open Settings → MCP.
2. Add a new MCP server.
3. Paste the Dwizi MCP JSON configuration.
4. Save.

Cursor will immediately discover all available tools.

Claude Desktop

1. Open Settings → Model Context Protocol.
2. Add a custom MCP server.
3. Paste the Dwizi MCP JSON configuration.
4. Save and restart if prompted.

Claude can now invoke Dwizi tools during conversations.

Continue.dev

1. Open ~/.continue/config.json.
2. Add a new MCP server entry using the Dwizi JSON config.
3. Save the file.
4. Restart the Continue extension.

Dwizi tools appear as callable actions.

Cline (VS Code)

1. Open the MCP Servers panel.
2. Add a new server using the Dwizi JSON configuration.
3. Save.
4. Reload the workspace.

All tools in the connected organization become available.

Automation & Workflow Platforms (HTTP)

Dwizi tools are also exposed as HTTP endpoints, making them compatible with no-code and low-code platforms.

n8n

1. Add an HTTP Request node.
2. Set:
   • Method: POST
   • URL: Dwizi tool URL
3. Enable JSON body.
4. Provide input payload.
5. Include authentication token if the tool is private.

Dwizi tools can now be used in workflows, triggers, and scheduled jobs.

Make (Integromat)

1. Add an HTTP → Make a request module.
2. Set method to POST.
3. Use the Dwizi tool URL.
4. Provide JSON input.
5. Add token if required.

Zapier

1. Use Webhooks by Zapier.
2. Choose POST.
3. Set the request URL to the Dwizi tool endpoint.
4. Provide JSON payload.
5. Add token for private tools.

Pipedream

1. Add an HTTP Request step.
2. Use POST.
3. Provide JSON input.
4. Include token if required.

Dwizi tools work naturally inside event-driven workflows.

API & Debugging

curl

curl -X POST https://tool.dwizi.com/your-tool-id \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "value": 42
    },
    "token": "YOUR_TOKEN_IF_REQUIRED"
  }'

Postman

1. Create a new POST request.
2. Set the tool URL.
3. Add header: Content-Type: application/json.
4. Provide JSON body.
5. Include token if private.

Useful for debugging, testing, and CI pipelines.

Execution Model

Client / Workflow → Interface (MCP / HTTP) → Dwizi Tool → Isolated Runtime

Dwizi ensures:

• Isolation per execution
• Deterministic inputs/outputs
• Secure lifecycle management

Security Notes

• Private tools require tokens
• Tokens can be rotated or revoked
• Public tools require no credentials
• No shared execution state

Next Steps

• Create your first tool
• Claude deep dive
• Security & access control

Copy-Paste Configuration Snippets

MCP JSON (Generic)

{
  "name": "dwizi",
  "transport": {
    "type": "http",
    "url": "https://mcp.dwizi.com/YOUR_ORG_ID"
  },
  "auth": {
    "type": "bearer",
    "token": "YOUR_TOKEN"
  }
}

Use this config in Cursor, Claude Desktop, Continue, or Cline.

HTTP Tool Invocation (Generic)

{
  "input": {
    "example": "value"
  },
  "token": "YOUR_TOKEN_IF_REQUIRED"
}