SETUP-FOR-DUMMIES
A step‑by‑step guide to get ChatAI up and running, even if you’ve never done this before.

1. Prepare Your Minecraft Server

1. Download Paper (recommended) or Spigot:
   
   • Paper 1.20+: https://www.spigotmc.org/
     
2. Install:
   
   • Place paper.jar (or spigot.jar) in an empty folder (e.g. mc-server/).
     
     
   • Create a launcher script (start.sh on Linux/macOS or start.bat on Windows):
     
     java -Xms1G -Xmx2G -jar paper.jar --nogui
3. Run once to generate folders:
   
   ./start.sh # or double‑click start.bat
4. Agree to EULA:
   
   • Open eula.txt, change eula=false to eula=true, save.
     

2. Install LM Studio & Llama 3.2 Instruct

1. Download LM Studio:
   
   • https://github.com/nomic-ai/LMStudio/releases
     
     
   • Choose the appropriate OS installer and follow instructions.
     
2. Obtain Llama 3.2 Instruct 3B model:
   
   • In LM Studio, go to Model Browser → search for Llama 3.2 Instruct → install.
     
3. Start LM Studio & load the model:
   
   • Launch LM Studio, open the Llama 3.2 Instruct model.
     
     
   • Note the API endpoint (default: http://127.0.0.1:1234/v1/chat/completions).
     

3. Set Up the Minecraft‑to‑LM Proxy
This small Node.js proxy translates Minecraft plugin requests into LM Studio calls.

1. Prerequisites: Node.js 18+ installed (https://nodejs.org/)
   
   
2. Create ``:
   
   // simple-proxy.js
   const express = require('express');
   const fetch = require('node-fetch');
   const app = express();
   
   app.use(express.json());
   
   const LM_HOST = 'http://127.0.0.1:1234';
   const LM_PATH = '/v1/chat/completions';
   
   const mcSchema = {
   type: 'object',
   properties: {
   answer: {
   type: 'string',
   description: 'The AI’s chat reply'
   }
   },
   required: ['answer']
   };
   
   app.post('/ai', async (req, res) => {
   const { model, system, messages } = req.body;
   if (!Array.isArray(messages) || !system || !model) {
   return res.status(400).json({ error: 'Bad request' });
   }
   const payload = { model, messages: [{ role:'system',content:system }, ...messages], response_format:{ type:'json_schema', json_schema:{ schema:mcSchema } } };
   try {
   const upstream = await fetch(`${LM_HOST}${LM_PATH}`, { method:'POST', headers:{'Content-Type':'application/json'}, body:JSON.stringify(payload) });
   if (!upstream.ok) return res.status(502).json({ error:`Upstream ${upstream.status}` });
   const data = await upstream.json();
   return res.json(data);
   } catch (err) {
   return res.status(500).json({ error: err.message });
   }
   });
   
   const PORT = 3004;
   app.listen(PORT, () => console.log(`AI‑proxy listening on http://localhost:${PORT}/ai`));
3. Install dependencies & run:
   
   npm init -y
   npm install express node-fetch
   node simple-proxy.js
4. Verify:
   
   • In your browser or via curl, POST a test JSON to http://localhost:3004/ai → should get a JSON { "answer": "…" }.
     

⚙️ 4. Configure the ChatAI Plugin

1. Place files:
   
   • Put ChatAI.jar in mc-server/plugins/
     
     
   • Create mc-server/plugins/ChatAI/
     
2. Copy stub configs: From the GitHub repo’s resources/ folder into plugins/ChatAI/:
   
   • config.yml, personalities.yml, names.yml, villagers.yml, memory.yml, villages.yml, missions.json
     
3. Edit ``:
   
   proxy-url: 'http://localhost:3004/ai'
   system-prompt: |
   You are a helpful Minecraft villager.
   model-name: 'llama-3.2-instruct'
   temperature: 0.7
   max-tokens: 4000
4. **Review **`` to tweak quests, rewards, counts.
   
   
5. (Optional) Examine AIClient.java in the plugin source for deeper customization (e.g. history length).
   

▶️ 5. Start Everything & Test

1. Run LM Studio (with Llama 3.2 Instruct).
   
   
2. Start the proxy: node simple-proxy.js.
   
   
3. Launch Minecraft server: ./start.sh.
   
   
4. In-Game:
   
   • Join the server, watch the welcome message.
     
     
   • Sneak + right-click a villager → chat UI appears.
     
     
   • Complete a quest: kill or fetch items → see action-bar progress.
     

Congratulations! ChatAI should now be fully functional.

If you hit any issues, double‑check log output in both the proxy console and the server console for error messages and missing configurations.