We've migrated from Express.js to Fastify for improved performance and efficiency. This change won't affect the API's functionality. Please report any issues you encounter.

This is a simple API for a character AI. The API provides endpoints for searching for characters, getting information about a character, and sending messages to a character AI.

• Character AI API
  • Table of Contents
  • Prerequisites
  • Installation
  • Running the API
  • API Endpoints
    • GET /api/search_character
    • GET /api/character_info
    • POST /api/send_message
  • Test
  • License
  • Acknowledgments
  • Live Demo

To run the API, you will need the following:

• Node.js
  • This project was built using Node.js version 20.0.0. You can check your version of Node.js by running node -v in your terminal.
• npm (Node Package Manager)
  • You can install these by following the instructions on the Node.js website.
• An access token for the character AI API
  • You can get this from the character AI service.

1. Clone the repository

2. Run npm install in the root directory to install the dependencies

   1. requires to install @xct007/chrt-ai

3. Create a .env file in the root directory with the following content or rename the .env.example file to .env and fill in the values:

   ACCESS_TOKEN=your_access_token
   ID_TOKEN=your_id_token
   PORT=8080

   • ACCESS_TOKEN: This is the access token for the character AI service.
   • ID_TOKEN: This is the id token for the character AI service.
   • PORT: This will set the port for the server to 8080. You can change this to any port you want. default is 3000 if not set.

Note: requires to install playwright dependencies. Run npx playwright install to install the dependencies.

To run the API, run npm start in the root directory. The API will be available at localhost:PORT.

Swagger documentation is available at /docs after running the API.

The API provides the following endpoints:

Returns a list of characters that match the search query. The search query should be sent as a query parameter in the URL.

curl -X GET \
  'http://localhost:8080/api/search_character?query=yor' \
  -H 'Content-Type: application/json'

{
	"status": true,
	"result": [
		{
			"status": false,
			"result": {
				"characters": [
					{
						"document_id": "3b1cb305-d8df-4fe7-80cf-b5de87e6d112",
						"external_id": "WsqG34NBsbCr3hxN7gJA_y5khYtVQzTD71IqdtfO57Y",
						"title": "A loving mom who's definitely not an assassin",
						"greeting": "Hi, I'm a loving mom who's definitely not an assassin.",
						"avatar_file_name": "uploaded/2022/11/28/aHHgTYRLA59ZMpLVfaRee_90iW42A_CeTvkkFlS3rOU.webp",
						"visibility": "PUBLIC",
						"participant__name": "Yor Forger",
						"participant__num_interactions": 58786141,
						"user__username": "CreativeUsername352",
						"priority": 0,
						"search_score": 71.66664
					}
				],
				"request_id": "t3q79PEARkuwthB_EddJsQ"
			}
		}
	]
}

Returns a list of all characters with their information. The character's external ID should be sent as a external_id parameter in the URL.

curl -X GET \
  'http://localhost:PORT/api/character_info?external_id=11' \
  -H 'accept: application/json'

{
	"status": true,
	"result": {
		"character": {
			"external_id": "WsqG34NBsbCr3hxN7gJA_y5khYtVQzTD71IqdtfO57Y",
			"title": "A loving mom who's definitely not an assassin",
			"name": "Yor Forger",
			"visibility": "PUBLIC",
			"greeting": "Hi, I'm a loving mom who's definitely not an assassin.",
			"avatar_file_name": "uploaded/2022/11/28/aHHgTYRLA59ZMpLVfaRee_90iW42A_CeTvkkFlS3rOU.webp"
		},
		"status": "OK"
	}
}

Sends a message to the character AI. The message should be sent in the body of the request as a JSON object with the following format:

{
	"external_id": "1234",
	"message": "Hello, world!"
}

The external_id is a unique identifier for the user sending the message. The message is the message to be sent to the character AI.

curl -X POST \
  http://localhost:PORT/api/send_message \
  -H 'Content-Type: application/json' \
  -d '{
    "external_id": "1234",
    "message": "Hello, world!"
}'

{
	"status": true,
	"result": {
		"replies": [
			{
				"text": "this is a reply",
				"uuid": "string",
				"id": 1
			}
		],
		"src_char": {
			"participant": {
				"name": "string"
			},
			"avatar_file_name": "string"
		},
		"is_final_chunk": true,
		"last_user_msg_id": 0,
		"last_user_msg_uuid": "xxx"
	}
}

To run the tests, run npm test in the root directory.

This project is licensed under the MIT License - see the LICENSE file for details.

This project was created as part of a technical interview process. Thank you for the opportunity to work on this project.

The API is Reverse-Engineered from the character AI service. Do not hope many features are available in this API.

If you have any questions or feedback, please feel free to reach out to me. I would love to hear from you!

A live demo of the API is available at https://apigratis.site/docs