What is tradingview-mcp?

TradingView MCP Bridge lets an AI assistant like Claude Code see and control your own TradingView Desktop app. It connects through the Chrome DevTools Protocol, a standard debugging interface built into Chromium and Electron apps, so the AI can read your charts and interact with them the same way a person would with a mouse and keyboard. Everything runs locally on your own machine. The tool does not talk to TradingView's servers, does not bypass any paywall, and does not store or send your market data anywhere. You still need your own paid TradingView subscription and the Desktop app installed, since this is only a bridge, not a replacement for TradingView itself. The author is upfront that it relies on undocumented internal TradingView interfaces, so an update to TradingView could break it without warning. Once connected, the AI can help write and debug Pine Script, the scripting language TradingView uses for custom indicators and strategies. It can also navigate charts by changing symbols and timeframes, add or remove indicators, draw trend lines and annotations, manage price alerts, step through historical price data to practice trades, take screenshots for visual analysis, and arrange multiple chart panes at once. Every one of these actions is also available as a command line tool called tv, so scripts can pipe its JSON output into other tools. Setup involves cloning the repository, running npm install, launching TradingView Desktop with a debug flag enabled, and adding the tool to Claude Code's configuration file. An optional Redis cache can speed up repeated chart reads, but the tool works fine without it. The README frames this as a research project studying how AI agents can work with financial software interfaces, and it explicitly states it is not a trading bot and does not place real trades. The README does not state a license for the project.