Fully Automated Twitter Bot

• Twitter: @BentleySuperior

Dependencies

---

• boto3 -
• gspread - Google API wrapper that allows for interaction with Google Sheets
• tweepy - Twitter API wrapper that allows for Interaction with Twitter

Technologies Utilized

---

• Python 3.9
• AWS ECR
  • Repo: twitter_bot_dkr_container
• AWS Lambda
  • Function: create_tweet_automate_cntr
• AWS EventBridge
  • Rule: multiplePostsTwitterBot
• AWS S3
  • Bucket: twitter-bot-cloudformation
• Docker Container
• Google Sheets API
• Twitter API

The Script - app.py

---

Problems Experienced

---

• BIGGEST PROBLEM: Dependencies Too Big. 264MB. Lambda Functions only upload zip files up to 50MB. Zip Files to S3 up to 250MB. Tried everything. 1 layer for all dependencies, 1 layer per dependency. Nothing worked and I was forced to go the Docker container route.
• M1 Macbook Architecture - x86/84
• Figuring Out How To Pull Google Credentials From S3 Bucket

Understanding The Dockerfile - Dockerfile

---

Building The Bot Docker Image And Storing It On AWS ECR - Command Line & AWS ECR

---

Build

• docker build -t ect-container-name:latest .

Tag

• docker tag twitter_bot_lambda_func_container:latest 580708967395.dkr.ecr.us-east-2.amazonaws.com/twitter_bot_dkr_container:latest

Push(To AWS ECR)

• docker push 580708967395.dkr.ecr.us-east-2.amazonaws.com/twitter_bot_dkr_container:latest

---

Deploying The Docker Image In An AWS Lambda Function - Command Line/AWS Lambda

---

Automating Tweet Creation Via AWS EventBridge Scheduler - AWS EventBridge

---

What's Next?

---

• Updating Image via CI/CD(Github Actions) every time the master branch is updated
• Convert to OOP