Deploy API Tools with Cloudflare Workers

Getting Started

Since the Dify API Extension requires a publicly accessible internet address as an API Endpoint, we need to deploy our API extension to a public internet address. Here, we use Cloudflare Workers for deploying our API extension.

We clone the Example GitHub Repository, which contains a simple API extension. We can modify this as a base.

git clone https://github.com/crazywoola/dify-extension-workers.git
cp wrangler.toml.example wrangler.toml

Open the wrangler.toml file, and modify name and compatibility_date to your application's name and compatibility date.

An important configuration here is the TOKEN in vars, which you will need to provide when adding the API extension in Dify. For security reasons, it's recommended to use a random string as the Token. You should not write the Token directly in the source code but pass it via environment variables. Thus, do not commit your wrangler.toml to your code repository.

name = "dify-extension-example"
compatibility_date = "2023-01-01"

[vars]
TOKEN = "bananaiscool"

This API extension returns a random Breaking Bad quote. You can modify the logic of this API extension in src/index.ts. This example shows how to interact with a third-party API.

// ⬇️ implement your logic here ⬇️
// point === "app.external_data_tool.query"
// https://api.breakingbadquotes.xyz/v1/quotes
const count = params?.inputs?.count ?? 1;
const url = `https://api.breakingbadquotes.xyz/v1/quotes/${count}`;
const result = await fetch(url).then(res => res.text())
// ⬆️ implement your logic here ⬆️

This repository simplifies all configurations except for business logic. You can directly use npm commands to deploy your API extension.

npm install
npm run deploy

After successful deployment, you will get a public internet address, which you can add in Dify as an API Endpoint. Please note not to miss the endpoint path, which is specifically defined in src/index.ts.

Adding API Endpoint in Dify
Adding API Tool in the App edit page

Additionally, you can use the npm run dev command to deploy locally for testing.

npm install
npm run dev

Related output:

$ npm run dev
> dev
> wrangler dev src/index.ts

 ⛅️ wrangler 3.99.0
-------------------

Your worker has access to the following bindings:
- Vars:
  - TOKEN: "ban****ool"
⎔ Starting local server...
[wrangler:inf] Ready on http://localhost:58445

After this, you can use tools like Postman to debug the local interface.

Other Logic TL;DR

About Bearer Auth

import { bearerAuth } from "hono/bearer-auth";

(c, next) => {
    const auth = bearerAuth({ token: c.env.TOKEN });
    return auth(c, next);
},

Our Bearer authentication logic is as shown above. We use the hono/bearer-auth package for Bearer authentication. You can use c.env.TOKEN in src/index.ts to get the Token.

About Parameter Validation

import { z } from "zod";
import { zValidator } from "@hono/zod-validator";

const schema = z.object({
  point: z.union([
    z.literal("ping"),
    z.literal("app.external_data_tool.query"),
  ]), // Restricts 'point' to two specific values
  params: z
    .object({
      app_id: z.string().optional(),
      tool_variable: z.string().optional(),
      inputs: z.record(z.any()).optional(),
      query: z.any().optional(),  // string or null
    })
    .optional(),
});

We use zod to define the types of parameters. You can use zValidator in src/index.ts for parameter validation. Get validated parameters through const { point, params } = c.req.valid("json");. Our point has only two values, so we use z.union for definition. params is an optional parameter, defined with z.optional. It includes a inputs parameter, a Record<string, any> type representing an object with string keys and any values. This type can represent any object. You can get the count parameter in src/index.ts using params?.inputs?.count.

Accessing Logs of Cloudflare Workers

wrangler tail

Reference Content

PreviousExternal Data ToolNextModeration

Last updated