Launch an Eliza Agent
Last updated
Was this helpful?
Last updated
Was this helpful?
Phala Cloud has introduced a new feature that allows users to deploy an without any coding. You can access this feature through . In this tutorial, we'll guide you through the process of deploying an agent named Trump, capable of tweeting in his distinctive tone.
Before you begin, it would be ideal to understand what an is. However, if you're unfamiliar with it, that's perfectly fine. You can start with a template and customize it to suit your needs. There are only two things you need to prepare:
An X account for the agent to connect and tweet
An OpenAI account for the agent to send requests and harness its intelligence
You can either upload your own character file or select one from the built-in options available in the . In this example, we'll choose Trump. Don't worry—you can still customize the template later if needed.
You can select multiple clients and configure the account information for each. In this example, we'll choose an X account and set the necessary information. You can also override the account information later if needed
Set basic information
Set TWITTER_USERNAME , TWITTER_PASSWORD , TWITTER_EMAIL environments that you set for your account.
Set TWITTER_2FA_SECRET
You should enable 2FA on your X account settings page and select an authentication app as the method. During setup, make sure to copy the secure code (it's 16 uppercase letters and numbers) by click the button under the QR code. This secure code will be the value you set for TWITTER_2FA_SECRET.
Others
To enable the agent to reply to messages more quickly, it is recommended to set TWITTER_POLL_INTERVAL=10, ``but means more quote usage from LLM provders.
In this step, you can review your configuration settings. You'll have the opportunity to update account information in the Environment Variables section and further customize the character file before deploying the agent.
Now, let’s check the agent X account, it already starting to mutter. And congratulation, you deployed an autonomous agent in TEE without a single line coding🎉.
💡 Join the technical support groups if you got any questions.
You need to choose a model provider for your agent to connect with and send requests to. In this example, we'll use , but you can also opt for other model providers like , which supports over 200 AI models with a unified account.
If you're unsure about other settings, you only need to configure the API key and API URL and model names. You can generate an API key using your OpenAI account on the . Type API URL: https://api.openai.com/v1
After clicking the "Complete" button, you will be redirected to the Cloud dashboard, If not yet login, it will ask to login first and all your data will store locally; If you don’t have an account, you can register one with this . Please be patient, as it may take several minutes to complete the deployment. Once the status changes from "STARTING" to "RUNNING," it indicates that your agent is now live.
CN: EN:
