noslip_quant_bot
quant information provider bot
README
No Slip Quant AI Trading Bot & SaaS Platform
No Slip is a state-of-the-art AI-assisted quantitative trading bot, swap analysis tool, and SaaS platform for S&P 500 stocks and major cryptocurrencies (BTC, ETH, SOL).
It integrates:
- 6-Agent Consensus Suite: Dynamic roundtable debate and voting between Macro, Trend, Value, Whale, Mean-Reversion, and Freqtrade ClucMay agents.
- Prophet Forecasting Engine: Multi-resolution timeframes analyzing daily and monthly trends, seasonality, and optimal trade timing.
- Multi-Strategy Real-time Monitor: Runs RSI Reversion, MACD Crossover, BB Squeeze Breakout, Spot Arbitrage, and Kimchi Premium Arbitrage in parallel.
- Telegram Interactive Bot: Persistent daemon allowing real-time portfolio tracking, debate simulations, backtest competition leaderboards, and on-demand stock/crypto analysis.
- Model Context Protocol (MCP): Custom MCP server exposing trading tools directly into Claude Desktop.
- SaaS Web App: Next.js dashboard with integrated Toss Payments SDK and credit system.
Core System Features
1. Multi-Agent Consensus Roundtable
The core valuation and signaling stack uses a weighted consensus from six specialized agents:
- Macro Agent: Tracks bond yields, inflation, USD index, crude oil, and VIX.
- Trend Agent: Evaluates EMA crossovers and momentum.
- Value Agent: Inspects valuation metrics and relative strength.
- Whale Agent: Tracks large-volume breakouts and trade signals.
- Mean-Reversion Agent: Searches for overextended local tops/bottoms.
- ClucMay Agent (Freqtrade): Implements the famous ClucMay strategy.
The agent weights are continuously tuned using a Policy Gradient Reinforcement Learning (RL) feedback loop based on realized market outcomes.
- Relevant files:
services/trader/multi_agent_consensus.py,services/trader/daily_reinforcement.py
2. Prophet Trend & Seasonality Forecasting
Features high-performance Prophet prediction cadences (intraday 4h for cryptos, daily for stocks). It captures:
- Macro trend slope (daily rate of change).
- Weekly & monthly seasonality normalized as a percentage of total forecasted price ($yhat$).
- Nyquist checking to prevent short-term seasonality aliasing on coarse data rules.
- Relevant files:
services/trader/main.py,services/trader/predict_signal.py,services/trader/champion_prophet.py
3. Telegram Interactive Bot Daemon (/기능 & /portfolio)
Runs a background long-polling listener allowing users to query and control the bot:
/portfolio(/포트폴리오): Displays S&P 500 asset allocations, virtual active holdings, and latest champion models registry./analyze <symbol>: Resolves Korean or English stock/crypto names and replies with the 6-Agent voting consensus./debate <symbol>: Initiates a live debate between agents. Users reply with their thesis, and the AI agent panel responds using Gemini or rule-based counterarguments./competition: Compares the No Slip Quant strategy against Jesse, Hummingbot, and Freqtrade in a 60-day historical backtest tournament./website(/웹사이트): Spawns an NPX localtunnel background process mappinglocalhost:3000to a public URL with a bypass code, letting you view your Next.js app on a mobile device.- Relevant files:
services/trader/telegram_interactive_bot.py
4. Parallel Multi-Strategy & Arbitrage Scanner
The background monitor daemon watches Binance, Bybit, and Upbit to scan:
- RSI Reversion: Rebound at oversold thresholds (<25).
- MACD Crossover: Volatility breakout at MACD Golden Cross.
- Bollinger Band Squeeze: Price breakout following tight volatility bands.
- Spot Arbitrage: Price inefficiencies between Binance Spot and Bybit Spot.
- Kimchi Premium Arbitrage: Real-time USD/KRW exchange-rate-adjusted premium tracking between Upbit and Binance Spot.
- Relevant files:
services/trader/whale_pump_monitor.py
5. Monthly Parameter Optimization Pipeline
Automatically updates the config file on the 1st of every month using numpy-accelerated strategy simulators:
- Downloads and caches 1-minute historical klines.
- Runs a calendar grid search to find optimal trigger thresholds, Hold times (H), Stop-Loss (SL), and Take-Profit (TP) targets for each strategy.
- Automatically updates
services/trader/model_cache/whale_config.jsondynamically loaded by the daemon. - Relevant files:
services/trader/optimize_monthly_strategies.py
Claude Desktop MCP Server Integration
No Slip includes a Model Context Protocol (MCP) JSON-RPC server that exposes our quant tools directly to your Claude Desktop AI assistant.
MCP Tools Provided
analyze_ticker: Prompts the 6-Agent Consensus engine to run a real-time evaluation.run_league_tournament: Executes the 60-day backtest tournament leaderboard.
Installation / Registration
Add the following to your Claude Desktop configuration file (typically ~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"NoSlipQuant": {
"command": "/Users/sunghoon/.gemini/antigravity/scratch/no-slip-saas/services/trader/.venv/bin/python",
"args": [
"-u",
"/Users/sunghoon/.gemini/antigravity/scratch/no-slip-saas/services/trader/mcp_server.py"
]
}
}
}
Local Development & Setup
1. Install Node Dependencies (Next.js)
npm install
2. Create Python Environment & Install Core Packages
python3 -m venv services/trader/.venv
services/trader/.venv/bin/pip install -r services/trader/requirements.txt
3. Environment Variables Config (.env)
Create a .env file in the root directory. Required parameters:
TELEGRAM_BOT_TOKEN=your_telegram_bot_token
TELEGRAM_CHAT_ID=your_allowed_chat_ids (comma-separated)
GEMINI_API_KEY=your_google_ai_studio_key
NEXT_PUBLIC_TOSS_CLIENT_KEY=your_toss_payments_client_key
TOSS_SECRET_KEY=your_toss_payments_secret_key
PREDICTION_API_TOKEN=your_shared_api_secret_token
4. Running the Applications
- Web App Dashboard (Next.js Dev Server):
npm run dev - Telegram Long-Polling Daemon & Web Tunnel:
Managed via macOS launchd plist or run directly:
services/trader/.venv/bin/python services/trader/telegram_interactive_bot.py - Real-time Price Strategy & Arbitrage Monitor:
services/trader/.venv/bin/python services/trader/whale_pump_monitor.py
production Deployment (macOS Daemon Setup)
The background jobs are orchestrated via macOS launchd plist templates:
com.noslip.telegram.plist: Runs the Telegram interactive bot daemon.com.noslip.whale.plist: Runs the price/arbitrage monitor.com.noslip.daily.plist: Spawns the S&P 500 & Crypto daily reports (run_daily.sh) at 08:30 KST.com.noslip.monthly.plist: Runs the monthly numpy backtest optimizer on the 1st of every month at 00:05 KST.
To load a daemon:
cp com.noslip.telegram.plist ~/Library/LaunchAgents/
launchctl load ~/Library/LaunchAgents/com.noslip.telegram.plist
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.