Skip to main content

🤖 People Trade AI — Built for People. Powered by AI.

Agent Name GitHub Maintainer Python Version Exchange Leverage Discord Community GitHub Architecture Post LinkedIn Launch Post

Democratizing Institutional Wyckoff Intelligence & High-Win Confluence Scalping for Everyone


🌟 Executive Overview

People Trade AI is a state-of-the-art, 24/7 autonomous algorithmic futures trading agent engineered specifically for CoinDCX Futures. Designed around Richard D. Wyckoff's institutional market cycle theory, the agent continuously scans 125+ high-volume altcoins, detecting smart money accumulation, manipulation springs, and mark-up expansions while enforcing strict 20X isolated leverage risk boundaries.

Equipped with a 24/7 Text-With-Agent REST console, a live 5-minute global NLP news sentiment engine, an on-chain Web3 whale liquidity tracker, and a real-time TradingView-integrated web terminal, People Trade AI operates completely silently without audio dependencies.


⚡ Core Institutional Features

🏛️ 1. Institutional Wyckoff Market Cycle Engine

  • Phase A (Stop Trend): Detects Preliminary Support (PS), Selling Climax (SC), and Automatic Rally (AR).
  • Phase B (Building Cause): Measures secondary tests (ST) and liquidity sweeps across resistance/support bounds.
  • Phase C (Smart Money Test): Pinpoints high-probability Accumulation Springs and Distribution Upthrusts (UTAD).
  • Phase D (Sign of Strength): Identifies Last Point of Support (LPS) and Breakouts above Ice Line.
  • Phase E (Markup Expansion): Rides momentum continuation with trailing stop-loss protection.

🛡️ 2. High-Win Signal Gate & Risk Auditor

  • Multi-Engine Confluence: Evaluates signals across 6 specialized engines (Wyckoff + Order Book + Math + Catalyst + Web3 + Risk Auditor).
  • Gate Threshold: Trades are executed ONLY when signal confluence meets or exceeds the 80% Win-Probability Threshold.
  • Full Market Asset Coverage: Multi-token trading engine scanning BTC, ETH, and 125+ top altcoins for maximum high-win signal opportunities.

💬 3. 24/7 Text-With-Agent Console

  • Interactive natural language interface accessible via HTTP REST (POST /api/agent/chat) or CLI (./chat.sh).
  • Instant diagnostic responses for queries like wyckoff, wallet, gate, news, pepe, sui, or start.

📰 4. Global 5-Minute News & Web3 Whale Tracker

  • 5-Min News Feed: Auto-refreshes top crypto catalysts every 5 minutes with VADER/NLP sentiment scoring (-1.0 to +1.0).
  • Web3 Whale Inflows: Tracks decentralized liquidity shifts and large wallet movements for explosive breakout setups.

🔒 5. Hardened Security & Manual-Only Withdrawals Policy

  • Automated Withdrawals Hard Disabled: The agent contains ZERO withdrawal capabilities (ALLOW_AUTOMATED_WITHDRAWALS = False).
  • Manual Control Only: Fund withdrawals can ONLY be initiated manually by the account owner inside the official CoinDCX App/Web.
  • API Scope Safety: Read & Futures Trading permissions ONLY. Never enable Withdrawal permissions on your API key.

🏗️ System Architecture

flowchart TD
    subgraph Market_Inputs["Market Data & Inputs"]
        CDCX_MD["CoinDCX 1m/5m Candlestick Stream"]
        W3_WHALE["Web3 On-Chain Liquidity Engine"]
        NEWS_NLP["Global Crypto News (5m NLP Refresh)"]
        X_CATALYST["X (Twitter) Catalyst Sentiment Feed"]
        TV_WEBHOOK["TradingView Webhooks FastAPI"]
    end

    subgraph Core_Engines["Institutional Analysis & Signal Gating"]
        WYCKOFF["Wyckoff Structural Engine (Phases A-E)"]
        MATH_ENG["Pre-Breakout Math & Volatility Engine"]
        ORDERBOOK["Order Book Liquidity Imbalance Auditor"]
        RISK_AUDIT["Institutional Risk Auditor & Capital Gate"]
    end

    subgraph Signal_Gate["High-Win Signal Confluence Gate"]
        GATE["High-Win Confluence Gate (Win Probability >= 80%)"]
    end

    subgraph Execution["Autonomous Execution & Monitoring"]
        CDCX_EXEC["CoinDCX 20X Futures Order Execution"]
        REST_CHAT["24/7 Text-With-Agent Console (/api/agent/chat)"]
        WEB_DASH["Silent Web Terminal (http://127.0.0.1:5000)"]
        KILL_SWITCH["Instant Emergency Kill-Switch (.killswitch)"]
    end

    CDCX_MD --> WYCKOFF
    W3_WHALE --> MATH_ENG
    NEWS_NLP --> WYCKOFF
    X_CATALYST --> WYCKOFF
    TV_WEBHOOK --> WYCKOFF

    WYCKOFF --> GATE
    MATH_ENG --> GATE
    ORDERBOOK --> GATE
    RISK_AUDIT --> GATE

    GATE -->|Approved Signal| CDCX_EXEC
    GATE -->|Rejected Signal| REST_CHAT

    CDCX_EXEC --> WEB_DASH
    CDCX_EXEC --> REST_CHAT
    KILL_SWITCH -.->|Halt Execution| CDCX_EXEC

📁 Repository Structure

people-trade-ai/
├── dashboard_server.py         # Flask Web Dashboard & 24/7 Text-With-Agent REST Server
├── coindcx_master_agent.py     # Central Autonomous Execution Orchestrator
├── wyckoff_engine.py           # Institutional Wyckoff Structural Analysis Engine
├── risk_auditor.py             # High-Win Gate Auditor & Equity Risk Manager
├── position_manager.py         # 20X Isolated Futures Position Lifecycle Manager
│
├── coindcx_client.py           # CoinDCX API Client (Signed HTTP Requests & HMAC)
├── coindcx_futures_mapper.py   # Symbol Normalization & Contract Precision Mapper
├── strategy_engine.py          # Candle Pattern & Momentum Breakdown Analytics
├── math_engine.py              # ATR Volatility & Technical Indicator Calculations
├── orderbook_engine.py         # Order Book Imbalance & Depth Analysis
│
├── catalyst_engine.py          # News & Twitter Catalyst Sentiment Aggregator
├── news_client.py              # 5-Minute Global News Retriever & NLP Sentiment Engine
├── web3_model.py               # Web3 On-Chain Whale Inflow & Liquidity Tracker
├── scan_web3_coins.py          # Altcoin Momentum Scanner (125+ Tokens)
│
├── agent_chat.py               # Interactive CLI Chat Console Handler
├── chat.sh                     # Bash Launcher for CLI Text Agent Console
├── autotrade.sh                # Autonomous Bot Launcher Script
├── config.py                   # Master Trading Parameters & Risk Settings
│
├── requirements.txt            # Python Dependencies
├── .gitignore                  # Security Rules (Protects .env & Secrets)
└── README.md                   # System Architecture & Documentation

🛠️ Installation & Setup

1. System Requirements & Wallet Funding

  • OS: Linux / macOS / WSL2
  • Python: 3.9+
  • CoinDCX Account: Valid Futures API Key & Secret

[!IMPORTANT] CRITICAL WALLET FUNDING REQUIREMENT: To execute trades, you MUST deposit or transfer funds into your CoinDCX USDT-M Futures Wallet.

  • Deposit Currency: USDT only (INR or Spot wallet balances cannot be used directly for USDT-M Futures).
  • Wallet Transfer Steps: Open CoinDCX App/Web $\rightarrow$ Wallets $\rightarrow$ Transfer $\rightarrow$ Transfer USDT from Spot/Main Wallet $\rightarrow$ USDT Futures Wallet.
  • Trading Margin: The agent reads available equity directly from your USDT Futures Wallet. Without USDT in the Futures Wallet, order placement will be rejected for insufficient margin!

2. Clone Repository

git clone https://github.com/rammaruboina-rgb/people-trade-ai.git
cd people-trade-ai

3. Virtual Environment & Dependencies

# Create virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

4. Security Configuration (.env)

Create a .env file in the root directory (Never commit this file!):

COINDCX_API_KEY=your_coindcx_api_key_here
COINDCX_API_SECRET=your_coindcx_api_secret_here
MODE=LIVE

🚀 Usage & Operating Modes

Mode 1: Launch Silent Dashboard & Agent Server

Start the dashboard web server on http://127.0.0.1:5000:

python dashboard_server.py

Open http://127.0.0.1:5000 in your browser to view live TradingView charts, real-time trade tapes, console logs, and text with the agent.

Mode 2: Interactive Text Console (CLI)

Interact with the agent directly from the terminal:

./chat.sh

Supported Query Badges: wyckoff, wallet, gate, news, sui, pepe, status.

Mode 3: Targeted Single-Coin Mode

Run the agent focused exclusively on one coin (e.g. SUI):

python coindcx_master_agent.py --coin SUI

⚙️ Institutional Risk & Safety Parameters

All core risk parameters are configured in config.py:

Parameter Value Description
LEVERAGE 20X Fixed Isolated Futures Leverage
WIN_RATE_THRESHOLD 80% Minimum Confluence Score for Entry
PROFIT_TARGET_PCT +20.0% Primary Take-Profit Target (ROE)
STOP_LOSS_PCT -10.0% Hard Stop-Loss Boundary (ROE)
BREAKEVEN_PROFIT_PCT +5.0% Move Stop-Loss to Entry Price
SYMBOL_BLACKLIST [] Multi-Asset Coverage (BTC, ETH, Altcoins Enabled)

🛑 Emergency Kill-Switch

To immediately halt trading and close all open futures positions:

touch .killswitch

Or click the EMERGENCY KILL-SWITCH button inside the web dashboard.


💬 Community & Discord Support

Join our official live Discord community for strategy discussions, market updates, and trading support:

👉 Join People Trade AI Discord Server
Discord Link: https://discord.gg/Gt3ZAfMWy


👤 Author & Maintainer

Maintained by rammaruboina-rgb 🚀


⚠️ Risk Disclaimer

This repository contains automated algorithmic cryptocurrency trading software. Cryptocurrency futures trading involves significant financial risk. Always test configuration parameters thoroughly before deploying capital.

Metadata

Release files for people-trade-ai 1.0.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for people-trade-ai 1.0.0
File Size Uploaded
people_trade_ai-1.0.0.tar.gz 63.3 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for people-trade-ai 1.0.0
File Interpreter ABI Platform
people_trade_ai-1.0.0-py3-none-any.whl Python 3 none any Details

Total release size: 126.6 kB

Release files / people_trade_ai-1.0.0.tar.gz

Download URL people_trade_ai-1.0.0.tar.gz
Size 63.3 kB
Tags Source
SHA-256 checksum
How to use checksums
564afe97b5b64277ce09dcb3b7b3faef08b914ad9a9c2c46d753c2cbc04b3642
BLAKE2b-256 checksum
How to use checksums
2440762c56c501c34ee8118076badf6ad711b7c7ca99322175308fdbae237d44
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 20, 2026.

Transparency log

Release files / people_trade_ai-1.0.0-py3-none-any.whl

Download URL people_trade_ai-1.0.0-py3-none-any.whl
Size 63.3 kB
Tags Python 3
SHA-256 checksum
How to use checksums
55d9792122556846ae3a632d4d76131fef4b008016857c901d23b15d5307f576
BLAKE2b-256 checksum
How to use checksums
604d5068d066c093b8f86ff9fd9fa3846707ebae8bf64877f8759c5aa7a680ec
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Aug 20, 2026.

Transparency log

Release history Release notifications | RSS feed

This release

1.0.0 This release

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page