TradingView MCP Bridge

The TradingView MCP Bridge is designed for agent-forward trading, where LLM agents assist human traders. The toolset follows several key design principles to optimize for LLM performance:

• Granular Tooling: Instead of a single coarse tool, the project provides ~84 granular tools (e.g., separate tools for quote, OHLCV, indicator values, pine lines, pine labels, etc.). This allows agents to request precise, small payloads to save context.
• Compact-by-Default Output: To manage context window constraints, tools return compact data by default (e.g., using summary: true and study_filter). Verbose mode is opt-in.
• Failure Transparency: The tools surface raw data rather than pre-interpreting it, allowing the agent's reasoning trace to reveal when it has misread a chart or encountered stale data.
• Pine Script Iteration Loop: The toolset supports a specific development loop for Pine Script: pine_set_source $\rightarrow$ pine_smart_compile $\rightarrow$ pine_get_errors to enable agents to debug and iterate on scripts.

To generate a comprehensive performance report for a Pine Script strategy on TradingView, follow this multi-step workflow using the available MCP tools:

1. Data Gathering

Collect the following data points using these tools:

• data_get_strategy_results: For overall metrics (net profit, win rate, profit factor, etc.).
• data_get_trades: For the individual trade list (up to 20 trades).
• data_get_equity: For equity curve data points.
• chart_get_state: To identify the current symbol, timeframe, and active studies.
• symbol_info: To obtain symbol metadata for context.

2. Visual Capture

Capture visual evidence of the strategy performance:

• Use capture_screenshot with the region "chart" to capture the chart with the strategy overlay.
• Use capture_screenshot with the region "strategy_tester" to capture the Strategy Tester panel.

3. Analysis Framework

Analyze the gathered data based on these criteria:

• Key Metrics: Evaluate Net Profit, % return, Total Trades, Win Rate, Profit Factor (target > 1.5), Max Drawdown ($ and %), Average Trade, Sharpe Ratio, and Max Consecutive Losses.
• Trade Analysis: Identify the largest winner/loser, reward:risk ratio (average winner vs average loser), long vs short performance, and time in market.
• Equity Curve: Assess smoothness, upward slope, drawdown periods, and profit distribution (consistency vs front/back-loaded).

4. Report Generation

Format the final output into a structured report including a Summary, Key Metrics table, Strengths, Weaknesses, and actionable Recommendations.

When building or modifying a Pine Script indicator or strategy in TradingView, follow this iterative development cycle:

1. Understand the Goal: Define the script type (indicator, strategy, or library), logic (entry/exit, oscillator, etc.), and visual requirements (overlay vs. separate pane).
2. Pull Current Source: If modifying an existing script, use the pull command to sync the current code to your local environment.
3. Write the Script: Save the complete code to scripts/current.pine. Ensure it includes the //@version=6 header and proper declarations.
4. Push and Compile: Inject the code into the TradingView Pine Editor and trigger a compilation.
5. Fix Errors: If compilation fails, edit scripts/current.pine based on the reported line numbers and descriptions, then push again.
6. Verify: Use screenshot and data retrieval tools to confirm visual and performance accuracy.
7. Iterate: Repeat the process for any requested changes.