TradingView MCP

TradingView MCP

An AI trading copilot that connects to TradingView Desktop via CDP, offering 84 MCP tools, 32 slash commands, Telegram alerts, and Dhan broker bridge for automated trading assistance and analysis.

Category
访问服务器

README

TradingView MCP — AI Trading Copilot

Your personal AI trading assistant that sees your TradingView chart, understands your trading style, and coaches you through every decision.

84 MCP tools | 30 skills | 32 slash commands | Telegram alerts | Dhan broker bridge | trader profile


One-Click Install

Paste this into Claude Code and it handles everything:

Install the TradingView MCP server from https://github.com/frozo-ai/frozo-tradingview-mcp.git — clone it, run npm install, add it to my .mcp.json, launch TradingView with debug port, and verify with tv_health_check. Then run /tv-profile-setup to set up my trader profile.

Or for any AI assistant with MCP support (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "tradingview": {
      "command": "node",
      "args": ["src/server.js"],
      "cwd": "/path/to/frozo-tradingview-mcp"
    }
  }
}

Demo — Real Live Data from TradingView

Every number below is pulled live from a running TradingView chart via CDP:

Demo

Run it yourself: bash demo/claude-demo.sh (requires TradingView Desktop running)


How It Works

Claude Code / Desktop / Cursor
       |
       | MCP protocol (stdio)
       v
  TradingView MCP Server (Node.js, 84 tools)
       |
       | Chrome DevTools Protocol (localhost:9222)
       v
  TradingView Desktop (Electron)
       |
  +----+----+----+
  |         |         |
Trader    Telegram   Dhan Broker
Profile   Bot API    (MCP bridge)

[!WARNING] Not affiliated with TradingView Inc. or Anthropic. Interacts with your locally running TradingView Desktop via Chrome DevTools Protocol. Review the Disclaimer before use.

[!IMPORTANT] Requires a valid TradingView subscription. Does not bypass any access controls. All data stays on your machine.


What Makes This Different

Most trading tools give you data. This one gives you a trading coach that:

  • Knows YOUR trading rules and enforces them
  • Validates every trade against YOUR checklist before you enter
  • Watches for YOUR behavioral weaknesses (revenge trading, cutting winners)
  • Sizes positions using YOUR risk parameters with F&O lot size awareness
  • Runs YOUR pre-market routine automatically every morning
  • Sends alerts to YOUR phone via Telegram
  • Bridges directly to YOUR Dhan broker for execution
  • Reviews YOUR journal and finds patterns YOU can't see

Quick Start

1. Install

git clone https://github.com/frozo-ai/frozo-tradingview-mcp.git
cd frozo-tradingview-mcp
npm install

2. Launch TradingView with CDP

# Mac
./scripts/launch_tv_debug_mac.sh

# Windows
scripts\launch_tv_debug.bat

# Linux
./scripts/launch_tv_debug_linux.sh

# Or manually on any platform
/path/to/TradingView --remote-debugging-port=9222

3. Connect to Claude Code

Add to your project .mcp.json:

{
  "mcpServers": {
    "tradingview": {
      "command": "node",
      "args": ["/path/to/frozo-tradingview-mcp/src/server.js"],
      "cwd": "/path/to/frozo-tradingview-mcp"
    }
  }
}

4. Connect to Claude Desktop App

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (Mac):

{
  "mcpServers": {
    "tradingview": {
      "command": "node",
      "args": ["/path/to/frozo-tradingview-mcp/src/server.js"],
      "cwd": "/path/to/frozo-tradingview-mcp"
    }
  }
}

5. Verify

Ask Claude: "Use tv_health_check to verify TradingView is connected"

6. Set Up Your Profile

/tv-profile-setup

This guided interview builds trader-profile.json — your identity file that personalizes everything.


The 5-Phase Professional Workflow

Run /tv-pro-workflow and the AI auto-detects the current time and executes the right phase:

Phase 1: Pre-Market News (8:00 AM)

/tv-news

Fetches global market news via web search: US close, Asia futures, GIFT Nifty, India VIX, FII/DII data, RBI/Fed updates.

Phase 2: Morning Prep (8:30 AM)

/tv-morning-prep
  • GIFT Nifty gap estimate
  • India VIX threshold check
  • Expiry day detection (weekly/monthly Thursday)
  • Deep scan: Daily + trading TF with all indicators
  • Sector rotation heatmap (10 Indian sectors)
  • Structured morning brief with bias

Phase 3: Market Open (9:30 AM)

/tv-gap              # Classify gap type (Common/Breakaway/Exhaustion/Island)
/tv-opening-range    # IB analysis + open type (OD/OTD/ORR/OA) + draw levels
/tv-regime           # Market regime (Trending/Ranging/Volatile/Choppy)

Phase 4: Active Trading (9:30 - 3:15 PM)

/tv-validate-trade   # Go/no-go checklist against YOUR rules
/tv-position-size    # F&O-aware lot sizing (NIFTY 75, BANKNIFTY 30)
/tv-dhan-execute     # Place order on Dhan (with confirmation)
/tv-session          # Live dashboard: P&L, risk budget, trade count
/tv-divergence       # NIFTY vs BANKNIFTY/VIX correlation monitor

Phase 5: End of Day (3:15 PM)

/tv-eod-report       # Session summary + compliance audit
/tv-journal          # Auto-generate trade journal entry
/tv-journal-review   # Weekly behavioral pattern analysis

All 32 Slash Commands

Phase Command What It Does
Setup /tv-profile-setup Guided interview to build your trader profile
Setup /tv-layout Save/load multi-pane chart layouts
Setup /tv-telegram Set up Telegram bot for mobile alerts
Daily /tv-start-session One command: health check + crons + session tracker + morning prep
Daily /tv-pro-workflow Complete 5-phase professional workflow with news
Pre-Market /tv-news Global market news via web search
Pre-Market /tv-morning-prep Full morning briefing (GIFT Nifty, VIX, expiry, sectors)
Pre-Market /tv-sectors Sector rotation heatmap (10 Indian sectors)
Open /tv-gap Gap type classification
Open /tv-opening-range Initial Balance analysis + draw levels
Open /tv-regime Market regime detector
Analysis /tv-analyze-chart Full chart analysis report
Analysis /tv-multi-tf Multi-timeframe confluence (3-4 TFs, scored 1-10)
Analysis /tv-divergence Nifty vs BankNifty/VIX/USDINR divergence monitor
Analysis /tv-compare Compare current chart to a historical date visually
Analysis /tv-backtest-pattern Scan history for pattern occurrences + statistics
Trading /tv-validate-trade Go/no-go checklist with expiry + session awareness
Trading /tv-position-size F&O-aware position sizing (lot multiples)
Trading /tv-options Option strategy suggester (strikes from chart levels + VIX)
Trading /tv-dhan-execute TradingView analysis to Dhan broker execution bridge
Trading /tv-alert-analyze Instant GO/NO-GO when an alert fires (<30 sec)
Trading /tv-rules Natural language rules engine + monitoring
Trading /tv-similar-trades Find matching trades in your journal history
Trading /tv-session Live session dashboard (P&L, risk budget, time alerts)
Review /tv-journal Auto-generate trade journal entry
Review /tv-what-if "What if I held longer?" exit strategy comparison
Review /tv-eod-report End-of-day report with compliance audit
Review /tv-journal-review Weekly/monthly behavioral pattern analysis
Strategy /tv-strategy-report Backtest performance report
Strategy /tv-scan-symbols Multi-symbol screening
Development /tv-pine-generate Natural language to Pine Script indicator
Practice /tv-replay-coach Coached replay with behavioral tracking

Trader Profile

The trader-profile.json file defines WHO you are. Every skill adapts to it.

# Option 1: Guided interview
/tv-profile-setup

# Option 2: From template
cp trader-profile.example.json trader-profile.json
# Edit with your details

What the Profile Controls

Section What It Affects
identity Name, style (scalper/intraday/swing), experience level
instruments Primary symbols, watchlist, default timeframe
risk Account size, risk %, max daily loss, max consecutive losses, lot sizes
session Trading hours, no-trade zones, expiry day rules
entry_rules Required conditions, preferred conditions, forbidden conditions
exit_rules Stop method, target method, trailing rules, min R:R
indicators Trusted indicators, ignored indicators, custom signal definitions
behavioral_notes Known weaknesses, strengths, coaching tone preference
telegram Bot token, chat ID, which alerts to send
journal Directory, screenshot inclusion, emotion tracking

Telegram Alerts

Get trading notifications on your phone — trade entries, session warnings, morning briefs, EOD reports.

Setup (3 minutes)

  1. Message @BotFather on Telegram, send /newbot, get your bot token
  2. Message your bot, then visit https://api.telegram.org/bot<TOKEN>/getUpdates to get your chat_id
  3. Add to trader-profile.json:
"telegram": {
  "bot_token": "YOUR_TOKEN",
  "chat_id": "YOUR_CHAT_ID"
}
  1. Test: /tv-telegram or tv telegram test

What Gets Sent Automatically

Event Telegram Message
Morning prep completes Condensed brief with bias
Trade entry logged Entry details + size
Trade exit logged P&L result (WIN/LOSS)
2 consecutive losses Warning to stop
Session status RED Alert: limits hit
Alert fires GO/NO-GO analysis
EOD report Day summary with stats

Indian Market Features

Built-in support for NSE/BSE F&O trading:

  • Expiry detection: Auto-detects weekly (Thursday) and monthly (last Thursday) expiry
  • GIFT Nifty: Pre-market gap estimate from GIFT Nifty futures
  • India VIX thresholds: <13 low, 13-18 normal, 18-25 elevated, >25 high, >30 no-trade
  • F&O lot sizes: NIFTY (75), BANKNIFTY (30), FINNIFTY (40), MIDCPNIFTY (50), SENSEX (10)
  • Session time checks: Lunch hour avoidance, last 15 min warnings, first 5 min cooldown
  • Sector rotation: 10 NSE sectoral indices scanned and ranked
  • Option strategies: Strike selection based on chart levels + VIX

Dhan Broker Integration

If you have the dhan-tradehull MCP server connected, the bridge enables:

Chart Analysis (TradingView) --> Validation --> Position Sizing --> Order Placement (Dhan)
  • Setup validator says GO --> AI shows execution summary --> You confirm --> Order placed
  • Never auto-executes — always requires your explicit confirmation
  • Checks all session/risk/expiry rules before allowing execution

CLI

Every tool is also a tv CLI command with JSON output:

# Profile management
tv profile view              # Show your profile summary
tv profile init              # Create from template
tv profile validate          # Check for missing fields

# Risk management
tv risk position-size --entry 22700 --stop 22650  # Uses profile defaults

# Telegram
tv telegram test             # Test connection
tv telegram send -m "msg"    # Send message
tv telegram screenshot -f chart.png -c "Setup"

# Chart operations
tv status                    # Connection check
tv quote                     # Current price
tv ohlcv --summary           # Price summary
tv screenshot                # Capture chart

# All commands
tv --help

Scheduling

Remote (runs even without Claude Code open)

Morning news brief is pushed to Telegram at 8:27 AM IST via scheduled remote agent.

Local (runs during your Claude Code session)

Run /tv-start-session each morning — it sets up all intraday crons:

Time What Runs
9:33 AM Gap analysis + Opening Range
9:38 AM Market regime + Sector rotation
12:03 PM Mid-day dashboard + Divergence check
3:17 PM EOD report + Journal
Every 30 min Rules monitor

79 MCP Tools

Chart Reading

chart_get_state data_get_study_values quote_get data_get_ohlcv

Pine Script Graphics

data_get_pine_lines data_get_pine_labels data_get_pine_tables data_get_pine_boxes

Chart Control

chart_set_symbol chart_set_timeframe chart_set_type chart_manage_indicator chart_scroll_to_date chart_set_visible_range symbol_info symbol_search indicator_set_inputs indicator_toggle_visibility

Pine Script Development

pine_set_source pine_smart_compile pine_get_errors pine_get_console pine_get_source pine_save pine_new pine_open pine_list_scripts pine_analyze pine_check

Replay Mode

replay_start replay_step replay_autoplay replay_trade replay_status replay_stop

Risk Management

risk_position_size

Telegram

telegram_send telegram_send_screenshot telegram_trade_alert telegram_test telegram_status

Drawing & Alerts

draw_shape draw_list draw_remove_one draw_clear draw_get_properties alert_create alert_list alert_delete

Multi-Pane & Tabs

pane_list pane_set_layout pane_focus pane_set_symbol tab_list tab_new tab_close tab_switch

UI Automation

ui_click ui_keyboard ui_type_text ui_hover ui_mouse_click ui_scroll ui_find_element ui_evaluate ui_open_panel ui_fullscreen

Batch & Capture

batch_run capture_screenshot watchlist_get watchlist_add layout_list layout_switch depth_get

Health & Launch

tv_health_check tv_launch tv_discover tv_ui_state


30 Skills

Category Skill Purpose
Setup profile-setup Guided interview to build trader identity
Pre-Market morning-prep Full morning briefing with GIFT Nifty, VIX, expiry
Pre-Market sector-scan 10 Indian sector indices rotation heatmap
Open gap-analysis Gap type classification and fill probability
Open opening-range Initial Balance + open type classification
Analysis chart-analysis Full chart analysis report
Analysis multi-tf-confluence Multi-timeframe signal agreement scoring
Analysis regime-detector Trending/Ranging/Volatile/Choppy classification
Analysis divergence-monitor Nifty vs BankNifty/VIX/USDINR correlations
Analysis screenshot-compare Compare chart to historical date
Analysis pattern-backtest Scan history for pattern occurrences
Trading trade-setup-validator Go/no-go checklist with profile rules
Trading option-strategy Option strategy from chart bias + VIX
Trading dhan-bridge TradingView to Dhan broker execution
Trading alert-analyzer Instant analysis when alert fires
Trading rules-engine Natural language rules + monitoring
Trading trade-similarity Find similar trades in journal history
Trading session-tracker Live P&L dashboard and risk management
Review auto-journal Auto-generate trade journal entries
Review what-if-replay Exit strategy comparison
Review auto-report EOD report with compliance audit
Review journal-review Weekly/monthly behavioral pattern analysis
Strategy strategy-report Backtest performance report
Strategy multi-symbol-scan Batch symbol screening
Development pine-develop Pine Script write/compile/debug loop
Development pine-generator Natural language to Pine Script
Practice replay-practice Basic replay stepping
Practice replay-coach Coached replay with behavioral tracking
Presets layout-presets Save/load multi-pane configurations
Alerts telegram-alerts Telegram bot setup and notification management

Architecture

Claude Code / Desktop
       |
       | (stdio / MCP protocol)
       v
  MCP Server (Node.js)
       |
       | (Chrome DevTools Protocol, localhost:9222)
       v
  TradingView Desktop (Electron)
       |
  +----+----+
  |         |
Profile   Telegram     Dhan Broker
(.json)   (Bot API)    (MCP bridge)

Dependencies: @modelcontextprotocol/sdk, chrome-remote-interface (that's it)


Testing

npm test              # E2E + unit tests
npm run test:unit     # Pine analysis + CLI tests (no TradingView needed)
npm run test:all      # All tests including replay

29 tests covering: Pine Script static analysis, server-side compilation, CLI routing, replay mode, and input sanitization.


Research Context

This project explores: how can LLM agents interact with professional trading interfaces to support human decision-making?

Key findings:

  • Context management is the primary constraint (compact defaults are critical)
  • 79 tools don't confuse the agent with good instructions (CLAUDE.md decision tree)
  • Pine Script development is the strongest use case for agent assistance
  • Trader profile personalization dramatically improves recommendation quality
  • Behavioral coaching (revenge trade detection, rule adherence tracking) is uniquely valuable

See RESEARCH.md for open questions and findings.


Disclaimer

This project is provided for personal, educational, and research purposes only.

This tool uses Chrome DevTools Protocol (CDP), a standard debugging interface built into all Chromium/Electron applications. It does not reverse engineer any proprietary protocol, connect to TradingView's servers, or bypass any access controls. The debug port must be explicitly enabled by the user.

By using this software, you acknowledge:

  1. You are solely responsible for compliance with TradingView's Terms of Use and all applicable laws
  2. TradingView's Terms restrict automated data collection and scraping — this tool may conflict with those terms
  3. You assume all risk — the authors are not responsible for account bans, legal actions, or other consequences
  4. This tool must not be used for redistributing market data, circumventing subscriptions, automated trading using extracted data, or violating Pine Script authors' IP
  5. Market data remains subject to exchange and data provider licensing terms
  6. This tool accesses undocumented internal APIs that may break at any time

Use at your own risk. If unsure about compliance, do not use this tool.

Not affiliated with TradingView Inc. or Anthropic, PBC.

License

MIT — see LICENSE. Applies to source code only, not to TradingView's software, data, or trademarks.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选