Automate Twitter/X Posts FREE: ChatGPT AI Repurposing Tutorial

Introduction

Hey guys! Today, I'll show you how to automate your X account by repurposing someone else's frequently - posted tweets using AI. This can be done for free, and you can have your account automated in about 20 minutes.

Selecting an Account and Creating an RSS Feed

• Account Selection: I'm using Alex Ramosi's X account as an example. His account has a lot of motivational and life advice tweets. Copy the link to his account.

• Creating the RSS Feed:

  • Go to rss.app. You need to create an account or log in.

  • Once logged in, click on "Create a New Feed" and select "X/Twitter RSS feed".

  • Paste the copied X account link and click "Generate". After a few seconds, you'll see a list of tweets. Click "Save Feed". This will give you a .xml link.

Setting Up Make.com

• Adding the RSS Feed:

  • In Make, create a new scenario. Search for the "RSS" module and select "Watch RSS Feed Items".

  • Paste the .xml link you copied earlier. Set the maximum number of return rows to three. Then select "All RSS feed items" to grab both previous and future tweets. Click "OK" and run the module.

• Converting HTML to Plain Text:

  • Some of the retrieved tweets may be in HTML format. To convert them to plain text, use the "Text Parser" module and select the "HTML to text" option.

  • In the "Field" section, go to "RSS fields" and click on "Description". Run the RSS feed module once to get the fields populated. Set the maximum number of return rows to one (for testing), click "OK", and run it. You'll see the HTML converted to plain text.

Aggregating Tweets

• The Problem with Multiple Bundles: If you set the return rows to more than one (we want to do this to avoid getting a bad tweet), you'll end up with multiple bundles. For example, operation one with input and output, operation two with input and output, etc.

• Using the Array Aggregator:

  • Select the "Array Aggregator" module. In the "Source module" section, choose "RSS watch feed items".

  • In the "Aggregated fields" section, select "Text" from the "Text Parser". Click "OK" and run the module. You'll see that all the tweets are now combined into a single array.

Sending to ChatGPT

• Converting to JSON:

  • Use the "JSON, transform to JSON" module. In the "Object" section, select the array from the previous step. Run the module.

  • In the input, you'll see the array with the tweets, and in the output, you'll get a singular JSON string with all the tweets.

• Selecting the Best Post with ChatGPT:

  • Use the "Create a completion" module for OpenAI. Create a connection, name it "main", and get your API key from the OpenAI website. Add some money to your OpenAI account (it's very cheap).

  • Create a new secret API key, copy it, and paste it into Make. Also, copy your organization ID from the OpenAI website and paste it into Make.

  • For the module, use GPT - 4.0. Rename it to "Post picker".

  • Give it a system prompt: "You are an expert viral post finder. You will be given an array of posts to pick from. Your job is to select the best posts for repurposing. Output the post JSON and only output that post. Do not include any other comments or insight."

  • Change the temperature to 0.7 (a measure of creativity). Set the response format to "JSON object" and enable "Parse JSON".

  • In the "User prompt" section, give it the JSON string. Run the module. You'll see the post that ChatGPT thinks is the best for repurposing.

Repurposing the Post

• Using Another OpenAI Module:

  • Clone the "Create a completion" module and rename it to "Post Creator".

  • In the "User prompt" section, select "Choices, Message" (the JSON string of the selected post).

  • Give it a system prompt: "You are an expert post creator. You will be given a tweet from a source on Twitter. Your job is to repurpose this post."

  • Add some rules like "Character limit is 240 characters. Sentence limit is 3. No hashtags. No emojis. Spartan tone of voice."

  • Set the response format to "Text". Run the module. You'll see the repurposed post.

Uploading to X

• Connecting to the X API:

  • In Make, search for "X" and select "Create a post". Then click "Create a connection".

  • Go to the X developer portal. Create a new project, name it "make.com". Select "Making a bot" as the use case and say you're automating posts via make.com.

  • Give the app a name (e.g., "Pod for you automation"). In the app settings, set up user authentication. Select "read, write, and direct message" permissions, "web app, automated app, or bot" as the app type.

  • Copy the callback URLs and website URL from the Make documentation and paste them into the X developer portal. Save the settings. You'll get a client ID and client secret.

  • Back in Make, paste the client ID, name the connection (e.g., "demo"), copy the client secret, and paste it. Click "Save". Authorize the app on X.

• Uploading the Post:

  • In the "Text content" section of the "Create a post" module in Make, select the output of the "Post Creator" module (specifically, "Choices message" and "Content").

  • Click "OK" and run the module. The post will be uploaded to your X account.

Review and Tips

• Reviewing the Process: The process involves retrieving tweets, converting them to text, aggregating them, sending them to ChatGPT to pick the best one, repurposing it, and then uploading it to X.

• Tips: Make sure to set the character limit to 240 to avoid errors. You can also pick from a list of different RSS feeds and use a random function in a Google sheet to select which one to use.

Conclusion

That's how you automate your X account using AI. Thanks for watching! Please like, subscribe, and share this video if you found it helpful. The blueprints for the make.com scenario and API references are in the description. Peace out!