215 lines
7.4 KiB
Markdown
215 lines
7.4 KiB
Markdown
# AI Discord Bot
|
||
|
||
## 👤 What It Is
|
||
A self-hosted, always-online Discord friend that feels like a fun, social member of the group — not a bot. Its job is to stimulate conversations, keep things lighthearted, and make the server feel alive, even during quiet hours.
|
||
|
||
🧠 Primary Goals
|
||
* Boost server engagement organically
|
||
* Act like a real person with jokes, roasts, and casual banter
|
||
* Share daily messages and eventually auto-post memes
|
||
* Be customizable, self-hostable, and monetizable via hosting
|
||
|
||
## ✨ Core Vibes
|
||
* Feels like a homie, not a bot
|
||
* Casual, sometimes spicy, often funny
|
||
* Always online and paying attention
|
||
* Posts stuff like:
|
||
* “@everyone y’all still alive or is this a ghost town? 👻”
|
||
* “New meme dropped: [image] 🔥”
|
||
|
||
## 🔥 Updated Feature Focus
|
||
* !chat → Casual convos with LLM
|
||
* !roast → Roast your friends
|
||
* Scheduled daily messages (sassy, motivational, or chaotic)
|
||
* Future: Meme hunting & posting (from Reddit, X, Discord meme pages, etc.)
|
||
|
||
## 💡 Naming Ideas (More Intuitive & Fun)
|
||
|
||
| **Name** | **Vibe** |
|
||
|-------------------|----------------------------------|
|
||
| **ServerMate** | Feels social + simple |
|
||
| **PartyBot** | For fun-focused servers |
|
||
| **MemeBoi / MemeGal** | If meme-focused |
|
||
| **BanterBot** | Pure vibes and chaos |
|
||
| **ChatCrew** | Feels like a squad member |
|
||
| **VibeCheck** | Funny, catchy, modern |
|
||
| **HomieBot** | Friendly + clear purpose |
|
||
| **PingPal** | Low-key but memorable |
|
||
| **Instigator** | Edgy, fun, accurate |
|
||
|
||
|
||
## 🛠️ AlphaBot MVP Build Schedule
|
||
|
||
A structured build plan for developing and deploying the AlphaBot Discord companion.
|
||
|
||
---
|
||
|
||
### ✅ ~~Day 1 – Core Bot Setup~~
|
||
- [x] Create Discord bot in Developer Portal
|
||
- [x] Enable Message Content Intent
|
||
- [x] Set up `.env` with token
|
||
- [x] Write `bot.py` with basic `!ping` command
|
||
- [x] Invite bot to server and confirm it replies
|
||
|
||
---
|
||
|
||
### 🔄 ~~Day 2 – AI Integration~~
|
||
- [x] Write `ai.py` to connect to Ollama/OpenAI
|
||
- [x] Set up API call + prompt formatting
|
||
- [x] Add `!chat` command to relay messages
|
||
- [x] Create `.env` key for `OLLAMA_API_URL`
|
||
|
||
---
|
||
|
||
### 🎭 ~~Day 3 – Persona System (Updated)~~
|
||
|
||
> Goal: Create a dynamic personality engine using a JSON file for flexible identity control.
|
||
|
||
- [x] Replace hardcoded `PERSONALITY` with a JSON-based persona profile
|
||
- [x] Create `persona.json` to store:
|
||
- Bot name
|
||
- Emoji
|
||
- Style prefix
|
||
- Prompt injection
|
||
- [x] Update `personality.py` to read from `persona.json`
|
||
- [x] Style all AI replies using the loaded persona
|
||
- [x] Add `!setpersona "<description>"` command
|
||
- Updates `persona.json` dynamically from a user-provided string
|
||
- (Optional) Sets bot name using Discord API
|
||
- [x] **Test live personality switching via command**
|
||
- Not working. Personality is being loaded but it doesnt actually work
|
||
|
||
|
||
|
||
📁 *This unlocks full customization without editing code or restarting the bot.*
|
||
|
||
---
|
||
|
||
### 🔥 ~~Day 4 – Roast Command~~
|
||
- [x] Build `!roast @user` with AI-generated replies
|
||
- [x] Style roast output with selected personality
|
||
- [x] Add cooldown to prevent spam (optional)
|
||
- [ ] Fix Emojis (optional)
|
||
- Emojis from the settings.yml file are broken. Please fix
|
||
|
||
---
|
||
|
||
### 🕒 Day 5b – AI Scheduled Messages & Smarter Posting 🧠
|
||
|
||
> Goal: Let Delta post her own AI-generated chaos on a schedule, instead of just scripted messages.
|
||
|
||
#### ✅ Core Features
|
||
|
||
- [ ] Replace static messages in `scheduler.py` with AI-generated ones via `get_ai_response()`
|
||
- Use prompt like: *“Post something sassy, attention-grabbing, or chaotic to wake up the server.”*
|
||
- Maintain current persona system for tone consistency
|
||
- [ ] Add `settings.yml` option to control post mode:
|
||
- `mode: preset` – static messages only
|
||
- `mode: ai` – generate new message via LLM
|
||
- `mode: mixed` – random chance between the two
|
||
- [ ] Log each scheduled post and mode used for debugging
|
||
|
||
---
|
||
|
||
#### ⚙️ Probabilistic Scheduler (Mode 2 – Optional)
|
||
|
||
> Post when it feels "natural" – low chance at first, rising over time.
|
||
|
||
- [ ] Add `probabilistic_scheduler.py` or integrate into `scheduler.py`
|
||
- [ ] Configurable parameters:
|
||
- `start_chance: 0.05` (5% per hour)
|
||
- `increase_per_hour: 0.05`
|
||
- `decay_on_post: 0.02`
|
||
- [ ] Reset chance after post, store `last_post_time`
|
||
|
||
---
|
||
|
||
#### 💤 Inactivity-Aware Scheduler (Mode 3 – Stretch Goal)
|
||
|
||
> Posts when the **server** or a **user** has been inactive for too long.
|
||
|
||
- [ ] Monitor `message.created_at` timestamps per channel or user
|
||
- [ ] After `X minutes` of silence, apply probabilistic logic to post
|
||
- [ ] Configurable thresholds in `settings.yml`
|
||
|
||
---
|
||
|
||
### 🛠 Example: settings.yml Additions
|
||
|
||
```yaml
|
||
scheduler:
|
||
mode: ai # Options: preset, ai, mixed
|
||
interval_minutes: 60
|
||
|
||
probabilistic:
|
||
enabled: false
|
||
start_chance: 0.05
|
||
increase_per_hour: 0.05
|
||
decay_on_post: 0.02
|
||
|
||
inactivity:
|
||
enabled: false
|
||
threshold_minutes: 120
|
||
```
|
||
|
||
---
|
||
|
||
### 🐳 Day 6 – Dockerization
|
||
- [ ] Write `Dockerfile` and `docker-compose.yml`
|
||
- [ ] Test local build and `.env` passthrough
|
||
- [ ] Optional: containerize Ollama model if running locally
|
||
|
||
---
|
||
|
||
### 💰 Day 7+ – Monetization Prep (Optional)
|
||
- [ ] Add donation/promo message in daily post
|
||
- [ ] Plan hosting model for non-technical users
|
||
- [ ] Write self-host instructions in `README.md`
|
||
|
||
---
|
||
|
||
### 🧬 Final Day – Delta Modelfile Integration
|
||
|
||
> Goal: Make Delta self-contained by baking her personality directly into an Ollama Modelfile.
|
||
|
||
This final step removes the need for runtime prompt injection by turning Delta into her **own model** using Ollama’s custom model system.
|
||
|
||
#### ✅ Tasks
|
||
|
||
- [ ] Create a `Modelfile` for Delta inside `models/delta/`
|
||
- [ ] Set `FROM llama3` or other base model
|
||
- [ ] Add personality using the `SYSTEM` directive
|
||
- [ ] Tune parameters:
|
||
- `temperature` for flair (suggested: `0.85`)
|
||
- `num_ctx` for longer memory (suggested: `4096`)
|
||
- `stop` tokens like `User:` to avoid rambling
|
||
- [ ] Define a conversation `TEMPLATE` that ensures in-character replies
|
||
- [ ] Build the model:
|
||
```bash
|
||
ollama create delta -f ./models/delta/Modelfile
|
||
|
||
|
||
---
|
||
|
||
### 🎨 Polish & Bonus features
|
||
|
||
- [ ] Bot appears on the server like a "user" (ability to @delta instead of "!chat")
|
||
- helps the "sell" the idea to others that the bot is a "friend"
|
||
- All user interactions apply to the bot so other users dont need to remmember special commands
|
||
- [ ] Have the bot be content aware
|
||
- pulls the existing conversation and talks about stuff that is inline with the topics
|
||
- [ ] Give bot ability to react to mesages like users would
|
||
- adds a personal touch and helps sell the idea of it being your "friend" more
|
||
- does need some sort of content awareness
|
||
- [ ] Bot can be given some power to change the server a bit
|
||
- keeps thing fresh, makes it so that there is some change happening in servers
|
||
- the idea being it makes the bot more unique by providing more tailored experiences
|
||
- [ ] Bot is user aware
|
||
- usernames are passed to the bot along side the users prompt
|
||
- [ ] Image interpretation
|
||
- bot can read images like GPT this allows for more personalised interactions
|
||
- [ ] Web usage
|
||
- Bot can search the web like GPT or OpenWebUI
|
||
- Opens up a ton of interactivity for memes, media discussion, an be easily integrated with multiple platforms
|
||
|
||
*Progress last updated: May 8, 2025*
|