GitMe - AI Git Summarizer Bot 🤖

GitMe is a Discord bot that watches your GitHub repositories. It tracks Commits, Pull Requests, and Issues, using LM Studio (Gemma) to:

1. Analyze technical commit messages.
2. Summarize changes into simple, layman terms locally.
3. Post beautiful, color-coded notifications to your Discord channel.

Features

• 🚀 Lightweight: Built with Node.js, Express, and Discord.js.
• 🧠 AI-Powered: Uses a local LLM via LM Studio to explain code changes without API costs.
• 🔔 Event Tracking:
  • Commits: AI summaries, branch info, and commit hashes.
  • Pull Requests: Notifications for Opened, Closed, Reopened, and Merged events.
  • Issues: Notifications for Opened, Closed, and Reopened issues.
• 🎨 Smart Embeds:
  • Dynamic Colors: Green for Features/Opens, Red for Fixes/Closures, Purple for Merges.
  • Rich Details: Includes author avatars, branch names, and truncated messages for readability.
• 🛡️ Error Handling: Robust error handling for local API connection issues.
• 📂 Modular Architecture: Professional structure with separate services, controllers, and config.

Prerequisites

• Node.js (v18 or higher)
• Ngrok (to expose your local server to GitHub)
• A Discord Account & Server
• LM Studio (to run the local LLM)

---

🛠️ Step-by-Step Setup

1. Clone & Install

git clone https://github.com/gubbysbyte/gitme.git
cd gitme
npm install

2. Configure Environment Variables

1. Rename .env.example to .env.
2. Open .env and fill in your keys:

DISCORD_TOKEN="your_bot_token_here"
DISCORD_CHANNEL_ID="your_channel_id_here"
LM_STUDIO_BASE_URL="http://localhost:1234/v1"
LM_STUDIO_MODEL="google/gemma-3-4b"
PORT=3000

(Make sure to keep the quotes around the values!)

Where to get the keys:

• DISCORD_TOKEN: Go to Discord Developer Portal -> New Application -> Bot -> Reset Token.
• DISCORD_CHANNEL_ID: In Discord app, go to User Settings -> Advanced -> Enable Developer Mode. Then right-click your target channel -> Copy ID.
• LM_STUDIO_BASE_URL: Usually http://localhost:1234/v1 (Default port for LM Studio).
• LM_STUDIO_MODEL: The identifier of the model loaded in LM Studio (e.g., google/gemma-3-4b).

3. Invite Bot to Server

1. In Discord Developer Portal -> OAuth2 -> URL Generator.
2. Select Scopes: bot.
3. Select Bot Permissions: View Channels, Send Messages, Embed Links.
4. Copy the generated URL, open it in a browser, and invite the bot to your server.

4. Start the Application

First, start your local server. This runs the bot and listens for webhooks.

npm start

5. Expose Local Server (Ngrok)

In a new terminal window, run ngrok to give your local server a public URL:

npx ngrok http 3000

Copy the Forwarding URL (e.g., https://abcdef123.ngrok-free.app).

⚠️ Important: On the free ngrok plan, this URL changes every time you restart ngrok.

6. Configure GitHub Webhook

1. Go to your GitHub Repository -> Settings -> Webhooks.
2. Click Add webhook.
3. Payload URL: Paste your ngrok URL and append /webhook at the end.
   • Example: https://abcdef123.ngrok-free.app/webhook
4. Content type: Select application/json.
5. Which events?: Select Let me select individual events and check: Pushes, Pull requests, Issues.
6. Click Add webhook.

---

🧪 Testing

1. Make a change to a file in your repository.
2. Commit and push:

   git add .
   git commit -m "feat: testing the new bot"
   git push

3. Check your Discord channel!
   • You should see a message with "AI Summary".

Manual Test Script

You can also simulate a webhook without pushing to GitHub:

node tests/test-manual-events.js push
node tests/test-manual-events.js issue
node tests/test-manual-events.js pr

AI Summary Test

Test if your LM Studio integration is working correctly:

npm run test:gemma

(Make sure LM Studio is running and the model is loaded!)

---

🐛 Troubleshooting

Error	Cause	Solution
TokenInvalid	Empty .env or bad token	Check .env file. Ensure values are in quotes "".
Missing Access	Bot not in channel/server	Invite bot to server. Give View Channel & Send Message permissions.
ECONNREFUSED	LM Studio server not running	Open LM Studio and ensure the Local Server is Started.
Model Not Found	Wrong model name in .env	Ensure LM_STUDIO_MODEL matches the identifier in LM Studio.
Webhook 404	Wrong URL in GitHub	Ensure Payload URL ends with /webhook (e.g., ...ngrok-free.app/webhook).

☁️ Deployment (Render + UptimeRobot)

Since Heroku is no longer free without a card, we use Render.com. To prevent the bot from sleeping after 15 minutes, use UptimeRobot.

Step 1: Deploy to Render

1. Push this code to GitHub.
2. Log in to Render.com.
3. New + -> Web Service.
4. Connect your repo.
5. Build Command: npm install
6. Start Command: npm start
7. Environment Variables: Add your keys (DISCORD_TOKEN, etc).
8. Click Create Web Service.

Step 2: Keep It Awake (Free 24/7)

1. Copy your new Render URL (e.g., https://gitme-bot.onrender.com).
2. Go to UptimeRobot.com (Free account).
3. Add New Monitor:
   • Type: HTTP(s).
   • Friendly Name: My GitMe Bot.
   • URL: Paste your Render URL.
   • Monitoring Interval: 5 minutes.
4. Click Create Monitor.

⚠️ IMPORTANT: Update Webhook

Update your GitHub Webhook URL to: https://gitme-bot.onrender.com/webhook

---

🤝 Contributing

Contributions are welcome! Please read our Contributing Guide for details on how to get started.

Please note that this project has a Code of Conduct. By participating, you agree to abide by its terms.

🔒 Security

To report a security vulnerability, please see our Security Policy.

📝 Changelog

See docs/CHANGELOG.md for a list of notable changes.

License

This project is licensed under the MIT License.