Descriptions:
This video demonstrates a fully automated pipeline for building and back-testing cryptocurrency trading strategies using Claude Code, Cursor AI, and a custom MCP server connected directly to TradingView. The presenter from Trading with DaviddTech uses Claude Opus 4.6 as the core agent, issuing a natural-language prompt that instructs it to select indicators, write Pine Script strategy code, deploy it to TradingView, and iterate through back-testing runs — without the user writing a single line of code.
The strategy targets Ethereum on a scalping timeframe and draws on indicators from TradingView developer BlackCat, including the CCI Trader, God Hunter Scalping indicator, and an ADX filter for volatility confirmation. A custom ‘David Tech boilerplate’ provides pre-built risk management scaffolding including multi-level take-profit targets and ATR band-based stop-loss placement. Claude runs the full optimization loop autonomously over roughly an hour before returning its final configuration.
The workflow showcases MCP as a practical bridge between an AI coding agent and a live trading platform, enabling non-programmers to produce and iterate on algorithmic strategies through conversational prompts alone. Back-testing results for the final Ethereum scalping strategy are reviewed live on screen, and the presenter promotes a free end-of-month workshop on algorithmic trading with AI bots.
📺 Source: Trading with DaviddTech · Published March 16, 2026
🏷️ Format: Hands On Build
