Twitter Bot

An automated Twitter/X bot that posts tech content 5 times a day — completely free using GitHub Actions. No server required.

---

What It Does

Every few hours, the bot wakes up, picks one of three content types at random, and posts it to your Twitter/X account:

• Latest headlines from TechCrunch, Hacker News, The Verge, Ars Technica, and Dev.to
• Tech quotes, tips, facts, and dev humor from a built-in template library
• AI-generated tech insights powered by Hugging Face

---

Prerequisites

You need accounts on three platforms before starting:

• GitHub — runs the bot for free
• Twitter/X — the account the bot will post to
• Hugging Face — provides the AI text generation

---

Setup Guide

Step 1 — Fork or clone this repository

Click the Fork button on GitHub to copy this repo to your account. That's the version you'll configure and run.

Step 2 — Get a Hugging Face API token

1. Sign up or log in at huggingface.co
2. Click your profile picture → Settings → Access Tokens
3. Click New token
4. Give it any name, select Read role, and make sure "Make calls to the serverless Inference API" is checked
5. Copy the token — it starts with hf_

Step 3 — Export your Twitter cookies

The bot logs into Twitter using browser cookies from your real session. This avoids bot detection.

1. Log into x.com in Chrome or Firefox
2. Install the Cookie-Editor browser extension
3. Click the Cookie-Editor icon in your toolbar
4. Click Export → Export as JSON
5. This copies the cookies to your clipboard — keep it ready for Step 4

Step 4 — Add secrets to GitHub

In your forked repository on GitHub:

1. Go to Settings → Secrets and variables → Actions
2. Click New repository secret and add each of the following:

Secret name	Value
TWITTER_USERNAME	Your Twitter/X username (without @)
TWITTER_PASSWORD	Your Twitter/X password
TWITTER_COOKIES	The full JSON you copied in Step 3
HUGGINGFACE_API_KEY	Your Hugging Face token from Step 2

Step 5 — Enable GitHub Actions

1. Go to the Actions tab in your repository
2. If prompted, click "I understand my workflows, go ahead and enable them"
3. The bot will now run automatically 5 times per day

To test it immediately, go to Actions → Post Tweet → Run workflow → Run workflow.

---

Posting Schedule

The bot runs at these times (UTC) every day:

Run	UTC time
Morning	8:00 AM
Midday	12:00 PM
Afternoon	4:00 PM
Evening	8:00 PM
Night	11:00 PM

To change the schedule, edit the cron expressions in .github/workflows/post.yml.

---

Customization

Add your own templates

Open data/templates.json and add entries following this format:

{
  "type": "tip",
  "content": "Your tweet text here\n\n#YourHashtag"
}

Types can be anything: quote, tip, fact, meme.

Add or remove RSS feeds

Open data/rss_feeds.json and add entries:

{
  "name": "Feed Name",
  "url": "https://example.com/feed.rss"
}

Any RSS or Atom feed URL works.

---

Running Locally

If you want to run the bot on your own machine:

1. Install Go 1.21+
2. Clone the repository
3. Copy the example env file and fill in your credentials:

cp .env.example .env

Edit .env:

TWITTER_USERNAME=your_username
TWITTER_PASSWORD=your_password
HUGGINGFACE_API_KEY=hf_your_token

4. Export your Twitter cookies to cookies.json in the project root (same process as Step 3 above)

5. Run the bot:

go run .

---

Troubleshooting

The workflow fails with "session invalid or expired" Your Twitter cookies have expired. Repeat Step 3 and update the TWITTER_COOKIES secret. Cookies typically last 30–90 days.

The workflow fails with "tweet composer not found" Twitter may have updated their page structure. Open an issue with the error and any screenshots from the artifacts.

The AI post fails Your Hugging Face token may have expired or lost its Inference API permission. Generate a new token and update the HUGGINGFACE_API_KEY secret.

I want to see what the browser sees When the workflow fails, GitHub automatically uploads debug screenshots as artifacts. Go to the failed Actions run → Summary → Artifacts → download debug-screenshots.

---

How It Works (technical)

1. GitHub Actions triggers the workflow on a cron schedule
2. The bot picks a random content type (template / RSS / AI)
3. For AI posts, it calls the Hugging Face Inference API
4. For RSS posts, it fetches the latest item from a random feed
5. It launches a headless Chromium browser, injects your session cookies, navigates to x.com, and posts the tweet
6. Screenshots are saved for debugging if anything goes wrong

---

License

MIT