# Kraken Discord Bot A standalone ComfyUI custom node that turns your ComfyUI instance into a Discord image generation bot. No complex workflows needed - just one node that handles everything! ![Workflow Screenshot](screenshots/workflow.png) ## Features - **All-in-One Node**: Single node handles Discord bot, image generation, and response - **Simple Setup**: Just add your Discord token, select a model, and queue - **Discord Commands**: Users can generate images with `!generate ` - **LoRA Support**: Up to 3 LoRAs with automatic CivitAI trigger word fetching - **Upscale Models**: Optional upscaling with RealESRGAN or other models - **AI Prompt Enhancement**: `--enhance` flag uses Qwen2-VL to expand simple prompts - **Style Presets**: Built-in styles (anime, photorealistic, fantasy, etc.) - **Quick Modifiers**: Instant quality/lighting/atmosphere tags (`--cinematic`, `--fog`, etc.) - **Size Presets**: Easy aspect ratios (`--portrait`, `--landscape`, `--wide`, etc.) - **Parameter Control**: Users can customize steps, CFG, size, seed via flags - **Rate Limiting**: Configurable per-user cooldown to prevent spam - **Queue Management**: Handles multiple requests with queue position feedback - **Persistent Config**: Settings are saved and restored between sessions ## Screenshots ### Discord Result ![Discord Result](screenshots/discord_result1.png) ### Help Menu ![Help Menu](screenshots/helpmenu.png) ## Installation ### Method 1: ComfyUI Manager 1. Open ComfyUI Manager 2. Search for "Kraken Discord Bot" 3. Click Install ### Method 2: Git Clone ```bash cd ComfyUI/custom_nodes git clone https://github.com/krakenunbound/kraken-discord-bot.git cd kraken-discord-bot pip install -r requirements.txt ``` ### Method 3: Download ZIP 1. Download this repository as ZIP 2. Extract to `ComfyUI/custom_nodes/kraken-discord-bot` 3. Install requirements: `pip install discord.py requests` ## Setup ### 1. Create a Discord Bot 1. Go to [Discord Developer Portal](https://discord.com/developers/applications) 2. Click "New Application" and give it a name 3. Go to "Bot" section and click "Add Bot" 4. Under "Privileged Gateway Intents", enable **Message Content Intent** 5. Copy the bot token (you'll need this) ### 2. Invite Bot to Your Server 1. Go to "OAuth2" > "URL Generator" 2. Select scopes: `bot` 3. Select permissions: `Send Messages`, `Embed Links`, `Attach Files`, `Read Message History` 4. Copy the generated URL and open it to invite the bot ### 3. Configure the Node 1. Add the "Kraken Discord Bot (All-in-One)" node to your workflow 2. Paste your Discord token in the `discord_token` field (shows "Token Loaded" once saved) 3. Select a checkpoint model 4. (Optional) Select up to 3 LoRAs - triggers are auto-fetched from CivitAI 5. (Optional) Select an upscale model for `--upscale` support 6. Configure default settings (steps, CFG, size) 7. Queue the workflow - the bot will start and wait for commands! ## Discord Commands ### Generate Images ``` !generate !gen !g ``` ### Parameter Flags ``` --negative What to avoid in the image --steps <1-50> Number of generation steps (default: 20) --cfg <1-20> Guidance scale (default: 7) --seed Specific seed for reproducibility --style Style preset (see below) --enhance Use AI to expand your prompt (slower but better) --upscale Upscale the result (if model configured) ``` ### Size Preset Flags ``` --square 1024x1024 (default) --portrait 832x1216 --landscape 1216x832 --wide 1344x768 --tall 768x1344 --phone 832x1216 (same as portrait) --desktop 1216x832 (same as landscape) ``` ### Quick Modifier Flags Add quality and style enhancements instantly: | Category | Flags | |----------|-------| | Quality | `--masterpiece` `--detailed` `--hd` `--professional` | | Lighting | `--cinematic` `--dramatic` `--soft` `--golden` `--neon` `--backlit` | | Atmosphere | `--fog` `--mist` `--rain` `--snow` | | Time | `--night` `--day` `--sunset` `--dawn` | | Style | `--epic` `--dark` `--vibrant` `--muted` `--vintage` `--noir` | | Camera | `--bokeh` `--dof` `--wideangle` `--macro` | | Artistic | `--painterly` `--ethereal` `--gritty` `--clean` | ### Style Presets Use with `--style `: - `photorealistic` - Realistic photo style - `anime` - Anime/manga style - `fantasy` - Epic fantasy art - `scifi` - Science fiction - `artistic` - Painterly style - `cinematic` - Movie-like shots - `cute` - Adorable kawaii style - `dark` - Dark and moody - `vintage` - Retro aesthetic - `minimalist` - Clean and simple ### Examples ``` !generate a majestic dragon flying over mountains --cinematic --fog --masterpiece !gen cyberpunk city at night --style scifi --neon --rain --wide !g anime girl with sword --style anime --portrait --detailed !generate cute cat playing --soft --vibrant --bokeh !generate epic battle scene --enhance --upscale ``` ### Other Commands ``` !help - Show help message with all options !status - Show bot status and queue ``` ## Node Settings | Setting | Description | Default | |---------|-------------|---------| | discord_token | Your Discord bot token | (required) | | checkpoint | Model to use for generation | (required) | | default_negative | Default negative prompt | ugly, blurry... | | default_steps | Default generation steps | 20 | | default_cfg | Default CFG scale | 7.0 | | default_width | Default image width | 1024 | | default_height | Default image height | 1024 | | sampler_name | Sampler to use | euler_ancestral | | scheduler | Scheduler to use | normal | | lora_1/2/3 | Optional LoRA models | None | | lora_X_strength | LoRA strength | 1.0 | | use_civitai_triggers | Auto-fetch trigger words | True | | upscale_model | Upscale model for --upscale | None | | rate_limit_seconds | Cooldown between requests per user | 30 | | max_queue_size | Maximum pending requests | 10 | | allowed_channels | Restrict to specific channel IDs | (all) | ## Troubleshooting ### Bot not responding - Make sure "Message Content Intent" is enabled in Discord Developer Portal - Check that the bot has permissions in the channel - Verify the token is correct ### Generation errors - Ensure you have a valid checkpoint selected - Check ComfyUI console for error messages - Make sure you have enough VRAM ### --enhance not working - Requires `transformers` and `accelerate` libraries - **Standard install:** `pip install transformers accelerate` - **Portable ComfyUI:** `.\python_embeded\python.exe -m pip install transformers accelerate` - First use downloads Qwen2-VL-2B (~4GB) - Needs additional VRAM (unloads after enhancement) ### --upscale not working - Select an upscale model in the node settings - Requires `spandrel` library - **Standard install:** `pip install spandrel` - **Portable ComfyUI:** `.\python_embeded\python.exe -m pip install spandrel` ## Requirements - ComfyUI (latest version recommended) - Python 3.10+ - discord.py >= 2.0 - requests - transformers + accelerate (optional, for --enhance) - spandrel (optional, for --upscale) - A Discord bot token - GPU with sufficient VRAM for your chosen model ## Changelog ### v1.1.0 (2024-12-09) - **LoRA Support**: Added 3 LoRA slots with automatic CivitAI trigger word fetching - **Upscale Models**: Added `--upscale` flag for on-demand upscaling - **AI Prompt Enhancement**: Added `--enhance` flag using Qwen2-VL-2B - **Size Preset Flags**: Direct flags like `--portrait`, `--landscape`, `--wide` - **Quick Modifiers**: 30+ modifier flags for instant quality/style boosts - **Improved Help**: Help command now uses Discord embed (prettier + no char limit) - **Bug Fixes**: Fixed parser bug where multiple boolean flags wouldn't parse correctly - **Single Token Field**: Simplified token input (shows "Token Loaded" when saved) ### v1.0.0 (2024-12-08) - Initial release - All-in-one Discord bot node - Basic generation with style presets - Rate limiting and queue management ## License MIT License - feel free to use, modify, and distribute! ## Credits - Built for ComfyUI - Uses discord.py for Discord integration - Qwen2-VL-2B for prompt enhancement - Kraken theme styling included --- Made with love for the AI art community!

Kraken Discord Bot Status

Inputs (0)

Outputs (1)