Cryto Trading AI Assistant
AI guardian for manual crypto traders — risk monitoring, strategy validation & emotional trading detection. No trade execution.
Install / Use
npx skills add aitradingbotspro/cryto-trading-ai-assistantInstalls into whichever agent you are using.
README
Strategy Guardian AI
AI-powered risk and strategy guardian for cryptocurrency traders.
Strategy Guardian AI monitors your trading in real time: it validates risk rules, detects emotional trading patterns, and enforces your strategy—without placing or closing any trades. Built for traders who want oversight and intelligence, not full automation.
I can help you grow your profits, tailor the system to what you want, and support you in making your trading safer and more rewarding—with clear rules, risk controls, and ongoing guidance.
Contact
I have years of real trading experience building and running automation like this in live markets. If you want help tailoring the bot to your strategy and improving your results, feel free to contact me on Telegram any time.
Screenshots
| | |
|---|---|
|
|
|
|
|
|
Overview
Strategy Guardian AI does not execute trades. It acts as a watchful layer around your activity:
| Capability | Description | |------------|-------------| | Trade Risk Monitor | Alerts when a trade exceeds your limits (max risk per trade, daily loss, leverage, position size). | | Strategy Rule Validator | Verifies entries and exits against your defined rules (e.g. RSI, support, trend). | | Emotional Trading Detector | Flags revenge trading, overtrading, size increases after losses, and off-hours activity. | | Market Risk Scanner | Warns before entry (volatility spikes, liquidation clusters, funding rates). | | AI Trade Feedback | Post-trade scores and suggestions (entry quality, risk management, strategy compliance). |
Alerts are delivered via Telegram or Discord and persisted for the dashboard.
Tech Stack
| Layer | Stack | |-------|--------| | Backend | Python 3.11+, FastAPI | | AI / ML | Scikit-learn (PyTorch optional) | | Data | SQLite (default), PostgreSQL, Redis | | Alerts | Telegram, Discord | | Dashboard | React, TypeScript, Vite |
Quick Start
# Clone and enter the repository
git clone <repository-url>
cd cryto-trading-ai-assistant
# Backend: create virtual environment and install dependencies
python -m venv .venv
.venv\Scripts\activate # Windows
# source .venv/bin/activate # Linux/macOS
pip install -r requirements.txt
# Configure environment (copy and edit)
copy .env.example .env
# Start the API server
python run.py
# Alternative: uvicorn app.main:app --reload
# Dashboard (optional; from repository root)
cd dashboard && npm install && npm run dev
# Open http://localhost:5173
Configuration
Copy .env.example to .env and configure:
- Exchange:
BINANCE_API_KEY,BINANCE_API_SECRET(or Bybit equivalents) - Alerts:
TELEGRAM_BOT_TOKEN,TELEGRAM_CHAT_ID,DISCORD_WEBHOOK_URL(optional) - Database:
DATABASE_URL,REDIS_URL(optional for local development)
Project Structure
cryto-trading-ai-assistant/
├── app/
│ ├── main.py # FastAPI application, API routes, webhook
│ ├── config.py # Application settings
│ ├── database.py # Async SQLite/PostgreSQL
│ ├── guardian.py # Guardian orchestration
│ ├── models.py # Pydantic models
│ └── db/
│ ├── models.py # SQLAlchemy (settings, alerts, trades)
│ └── repository.py # CRUD operations
├── dashboard/ # React dashboard (alerts, risk, trade history)
├── modules/
│ ├── exchange_connector/ # Binance, Bybit
│ ├── risk_engine/ # Position risk, leverage, drawdown
│ ├── strategy_validator/ # Rule engine
│ ├── emotion_detector/ # Behavior model
│ ├── ai_analysis/ # Trade feedback, market risk scanner
│ └── alerts/ # Telegram, Discord
├── images/ # Screenshots for documentation
├── requirements.txt
├── .env.example
└── README.md
Workflow
- You open or manage a trade manually on your exchange.
- The guardian receives the trade via the exchange API or webhook (
POST /webhook/trade). - The risk engine evaluates size, leverage, and daily PnL.
- The strategy validator checks entry/exit rules against your saved settings.
- The behavior detector looks for emotional or impulsive patterns.
- Warnings are sent to Telegram/Discord and stored for the dashboard.
Webhook Integration
Send new trade events to run the full guardian pipeline (e.g. from TradingView or your own scripts):
curl -X POST http://localhost:8000/webhook/trade \
-H "Content-Type: application/json" \
-d '{"exchange":"binance","symbol":"BTCUSDT","side":"long","size":0.01,"entry_price":65000,"leverage":5}'
Use this URL as the webhook target in TradingView alerts or when a position is opened.
Dashboard
The React dashboard provides:
- Alerts — Real-time feed of all guardian warnings.
- Risk & Account — Account balance/equity (Binance/Bybit) and editable risk rules (max risk %, daily loss %, leverage, position %).
- Trades — History of trades monitored by the guardian, with candlestick and equity views.
Data is persisted in SQLite by default (or DATABASE_URL if set).
API Reference
| Endpoint | Description |
|----------|-------------|
| GET /api/settings | Retrieve risk and strategy settings |
| PUT /api/settings | Update risk and/or strategy rules |
| GET /api/alerts | List stored alerts (for dashboard) |
| GET /api/trades | List stored trades |
| POST /webhook/trade | Ingest new trade (run guardian and persist) |
| GET /account/{exchange} | Account snapshot |
| GET /positions/{exchange} | Open positions |
Strategy Guardian AI — Protection and intelligence for manual traders, without full automation.
Related Skills
python-debugpy
385.5kDebug Python with pdb, breakpoint(), post-mortem inspection, and debugpy remote attach.
skill-creator
385.5kCreate, edit, audit, tidy, validate, or restructure AgentSkills and SKILL.md files.
imsg
385.5kUse the imsg CLI from OpenClaw agents for iMessage/SMS DMs, groups, replies, reactions, polls, watching, and private-API actions.
claude-opus-4-5-migration
140.6kMigrate prompts and code from Claude Sonnet 4.0, Sonnet 4.5, or Opus 4.1 to Opus 4.5
