Roast Builder Engine
kevadiyamalay80-code/roast-builder-engine
Gera rebates inteligentes para mensagens agressivas usando IA
Ver no GitHubGerador de respostas IA que transforma mensagens agressivas em críticas inteligentes e bem-humoradas. Funciona com análise contextual de sentimento e oferece múltiplos níveis de intensidade para criar retruques adequados ao tom da mensagem original.
Instalação
npx convez-ai kevadiyamalay80-code/roast-builder-engine1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Responder a trolls e cyberbullying em redes sociais
- Gerar respostas bem-humoradas em conversas casuais
- Criar retruques contextualizados com diferentes níveis de agressividade
Veredito

Ferramenta válida para tratamento leve de conflitos online, mas com limitações éticas claras. A aplicação prática é restrita a contextos sociais informais; não deve ser usada em ambiente profissional ou situações críticas.
README
A tradução para o português deste README ainda está na fila.
Instant Roast Engine: AI-Powered Comeback Generator for Social Media & Messaging
Transform Mean Messages into Witty Retorts with Artificial Intelligence
Tired of thinking of the perfect comeback three hours later? The Instant Roast Engine is your digital shield against online negativity. Paste any mean-spirited or sarcastic message, select your desired intensity level, and receive a clever, context-aware retort in under two seconds. Built as a free, open-source skill for Claude Code, this tool turns digital confrontation into a game of wit you always win.
Unlike traditional chat applications or generic AI responses, this engine understands nuance. It doesn't just generate text—it crafts personalized comebacks that match the tone, intent, and emotional weight of the original message. Whether you're dealing with a subtle dig from a coworker or a full-blown verbal assault in a gaming lobby, this tool has your back.
🎯 What Is the Instant Roast Engine?
The Instant Roast Engine is a specialized AI-powered text processing system designed to transform negative social messages into clever, humorous, or devastatingly effective comebacks. Think of it as a verbal judo master—it takes incoming aggression and redirects it with precision and style.
This repository contains the complete source code, API integration patterns, deployment scripts, and example configurations for running your own instance of the engine. It's designed to work seamlessly with Claude AI and OpenAI GPT models, but can be adapted for any large language model.
Core Philosophy: Every mean message is an opportunity to demonstrate intelligence, humor, and emotional control. We don't escalate—we elevate.
🧠 How It Works
The engine operates on a multi-layered analysis pipeline:
flowchart TD
A[User Input: Mean Message] --> B[Sentiment Analyzer]
A --> C[Intent Classifier]
A --> D[Toxicity Scorer]
B --> E[Context Engine]
C --> E
D --> E
E --> F[Intensity Selector]
F --> G{User Tier Choice}
G -->|Soft| H[Gentle Humor Generator]
G -->|Medium| I[Clever Wit Generator]
G -->|Heavy| J[Sharp Retort Generator]
G -->|Trainwreck| K[Absolute Destroyer Generator]
H --> L[Persona Filter]
I --> L
J --> L
K --> L
L --> M[Safety Check]
M --> N[Final Comeback Output]
Each stage refines the output to ensure the response is:
- Contextually accurate
- Tonally appropriate for your chosen tier
- Free from harmful or bannable language
- Legally and ethically sound
🚀 Key Features
Intelligent Response Tier System
Choose from four escalation levels that determine the intensity of your comeback:
| Tier | Intensity | Best For |
|---|---|---|
| Soft | Gentle Humor | Family group chats, polite disagreements |
| Medium | Clever Wit | Professional settings, social media |
| Heavy | Sharp Retort | Bullies, aggressive trolls |
| Trainwreck | Absolute Destroyer | Final word in a toxic argument |
Multilingual Support (2026 Edition)
Fully localized in 24 languages including English, Spanish, Mandarin, Arabic, Hindi, French, German, Japanese, Korean, Russian, Portuguese, Italian, Turkish, Vietnamese, Thai, Dutch, Polish, Swedish, Norwegian, Danish, Finnish, Greek, Czech, and Romanian. The engine detects the input language automatically and responds in the same language.
Responsive Web Interface
Built with a mobile-first design philosophy. The interface adapts seamlessly to any screen size, from smartwatches to 4K monitors. No app downloads required—works entirely in-browser.
24/7 Customer Support
Every installation includes access to our community forum and automated documentation bot. For troubleshooting, query example patterns, or emotional support after a particularly brutal roast session, the community has your back.
OpenAI API and Claude API Integration
The engine supports dual model backends:
OpenAI API: GPT-4 Turbo, GPT-4, GPT-3.5 Turbo
Claude API: Claude 3 Opus, Claude 3 Sonnet, Claude 3 Haiku
Switch between providers with a single configuration change. Each model brings unique strengths to the roasting table.
💻 Example Profile Configuration
Create a .env file in your project root:
# Core Settings
ENGINE_NAME=InstantRoastV2
DEFAULT_TIER=medium
MAX_RESPONSE_LENGTH=280
ENABLE_CONTEXT_AWARENESS=true
# API Keys
OPENAI_API_KEY=sk-your-key-here
CLAUDE_API_KEY=sk-ant-your-key-here
# Persona Settings
ROAST_STYLE=witty
AVOID_PERSONAL_ATTACKS=false
PROFESSIONAL_MODE=false
USE_POP_CULTURE_REFERENCES=true
# Safety Limits
TOXICITY_CAP=0.7
BANNED_WORDS_EXCEPTION=false
SAFETY_FILTER_LEVEL=moderate
# Multilingual
DEFAULT_LANGUAGE=auto-detect
FALLBACK_LANGUAGE=en
# UI Preferences
THEME=dark
ANIMATION_SPEED=normal
SOUND_EFFECTS=false
⌨️ Example Console Invocation
Run the engine from your terminal:
# Basic usage
python roast.py --message "Your code is terrible and you should feel bad" --tier heavy
# Output:
# "My code may be terrible, but at least it exists. Unlike your contribution to this project."
# With context and persona
python roast.py --message "nobody asked you" --tier soft --context "colleague disagreement" --persona "friendly"
# Output:
# "You're right, nobody asked me. But sometimes the best answers come from unasked questions!"
# Using Claude API
python roast.py --message "you're literally the worst" --tier trainwreck --backend claude
# Output:
# "I appreciate you setting the bar so low. It makes excellence look even better when I casually vault over it."
📱 OS Compatibility
| Operating System | Minimum Version | Status | Notes |
|---|---|---|---|
| ✅ Windows | 10 (build 19041+) | Fully Supported | PowerShell 5.1+ required |
| ✅ macOS | 11 Big Sur+ | Fully Supported | M1/M2/M3 native |
| ✅ Linux | Ubuntu 20.04+ | Fully Supported | All major distros |
| ✅ Android | 10+ | Via Termux | Limited GUI |
| ✅ iOS | 15+ | Via a-Shell | Experimental |
| ✅ ChromeOS | 100+ | Linux container | Stable |
| ⚠️ FreeBSD | 13+ | Community Support | Manual setup |
| ❌ Raspberry Pi OS | 10+ | Partial Support | Performance limitations |
🛡️ OpenAI API and Claude API Integration
OpenAI Configuration
import openai
from roast_engine import InstantRoast
engine = InstantRoast(
api_type='openai',
model='gpt-4-turbo',
temperature=0.75,
presence_penalty=0.4,
frequency_penalty=0.3,
top_p=0.9
)
response = engine.generate(
message="Your opinion is irrelevant to me",
tier="medium",
context="online debate"
)
Claude API Configuration
import anthropic
from roast_engine import InstantRoast
engine = InstantRoast(
api_type='claude',
model='claude-3-opus-20240229',
temperature=0.8,
max_tokens=150,
top_k=50
)
response = engine.generate(
message="You couldn't even fix a typo",
tier="heavy",
context="code review"
)
Dual Backend Strategy
For maximum reliability, implement a fallback chain:
engine = InstantRoast(
primary='claude',
fallback='openai',
retry_count=3,
timeout=5
)
📜 License
This project is licensed under the MIT License. See the LICENSE file for full details.
MIT License
Copyright (c) 2026
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files...
⚠️ Disclaimer
Important Legal and Ethical Notice
The Instant Roast Engine is designed for entertainment, education, and psychological resilience training purposes only. By using this software:
- You accept full responsibility for any comebacks generated and their social consequences.
- Do not use this tool to harass, threaten, bully, or incite violence against any individual or group.
- The developers are not liable for damaged relationships, workplace conflicts, lost friendships, or any other negative outcomes resulting from the use of generated comebacks.
- Always consider the context of your social situation. A "trainwreck" level response may seem funny in a gaming lobby but could end a professional relationship.
- This tool does not provide psychological advice. If you are experiencing persistent bullying or harassment, please seek support from appropriate professional resources or trusted individuals.
- The AI model may occasionally generate inappropriate content despite safety filters. Review all responses before sending them.
- Respect platform terms of service. Some social media platforms prohibit automated or AI-generated responses.
Remember: The best comeback is sometimes no comeback at all. Use this power wisely.
🌟 SEO Keywords
Comeback generator AI, AI roast engine, smart reply bot, witty retort generator, insult comeback maker, AI insult response, social media shield tool, conversation AI tool, verbal defense AI, automated witty responses, chat retort generator, AI for social situations, online bullying defense tool, argument AI assistant, clever response generator, mean message handler, AI communication coach, roasting AI tool, chatGPT comeback maker, Claude AI skill, OpenAI roast tool.
🔮 Future Roadmap (2026)
- Voice Mode: Real-time verbal comeback generation for live conversations
- Emotion Detection: Camera-based facial expression analysis to gauge the target's reaction
- Cultural Context Engine: Automatically adjusts for regional humor differences
- Group Chat Mode: Simultaneously processes multiple conversation threads
- Comeback History Analyzer: Tracks patterns in conversations and suggests strategic responses
- API Marketplace: Share and sell custom persona packs
🤝 Contributing
We welcome contributions from developers, writers, comedians, and linguists. Whether you want to add new roast templates, improve the NLP pipeline, or create documentation in your native language, every contribution matters.
Check out CONTRIBUTING.md for our code of conduct and submission guidelines.
📦 Quick Start
git clone https://github.com/your-username/instant-roast-engine.git
cd instant-roast-engine
pip install -r requirements.txt
cp .env.example .env
# Edit .env with your API keys
python roast.py --message "test message" --tier soft
Built with ❤️ for everyone who's ever stared at a screen, speechless, while someone got the last word. Not anymore.