About Pine Script v6 Documentation
Pine Script v6 Documentation is an MCP server for AI assistants and Pine Script developers. It bundles v6 references and provides function checks and prompt templates to ground code generation in documentation.
Key features
- 01Search and read Pine Script documentation
- 02Check valid v6 functions
- 03Browse documentation resources
- 04Analyze errors or explain functions with prompts
- 05Convert v5 code to v6 syntax
Use cases
- 01Verify function names and parameters before generating code
- 02Look up concepts such as repainting and the execution model
- 03Investigate Pine Script compilation errors
- 04Convert v5 code to v6 syntax
MCP support
The MCP server provides documentation lookup, function validation, resource access, and prompt tools to clients. It can run locally over stdio or be accessed through its hosted Streamable HTTP service.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://pinescript-mcp.fly.dev/mcpclaude mcp add --transport http paulieb89-pinescript https://pinescript-mcp.fly.dev/mcp{"mcpServers":{"paulieb89-pinescript":{"type":"http","url":"https://pinescript-mcp.fly.dev/mcp"}}}Open in Cursor{"servers":{"paulieb89-pinescript":{"type":"http","url":"https://pinescript-mcp.fly.dev/mcp"}}}Open in VS Code{"mcpServers":{"paulieb89-pinescript":{"type":"http","url":"https://pinescript-mcp.fly.dev/mcp"}}}{"mcpServers":{"paulieb89-pinescript":{"serverUrl":"https://pinescript-mcp.fly.dev/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://pinescript-mcp.fly.dev/mcphttps://pinescript-mcp.fly.dev/sseTools
10This tool takes no parameters.
Details
From the project READMEpinescript-mcp
MCP server providing Pine Script v6 documentation for AI assistants (Claude, etc.).
Enables AI to:
- Look up Pine Script functions and validate syntax
- Access official documentation for indicators, strategies, and visuals
- Understand Pine Script concepts (execution model, repainting, etc.)
- Generate correct v6 code with proper function references
Quick Start (stdio)
Works with Claude Code, Claude Desktop, Gemini CLI, and any MCP client that supports stdio:
{
"mcpServers": {
"pinescript-docs": {
"type": "stdio",
"command": "uvx",
"args": ["pinescript-mcp"]
}
}
}
Public Server (No Install Required)
No Python or uvx needed — connect directly to the hosted server.
Streamable HTTP — Claude Code, mcpServers with type field):
{
"mcpServers": {
"pinescript-docs": {
"type": "http",
"url": "https://pinescript-mcp.fly.dev/mcp"
}
}
}
Windsurf — uses its own shape (see Windsurf docs). Edit ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"pinescript-docs": {
"serverUrl": "https://pinescript-mcp.fly.dev/mcp"
}
}
}
ChatGPT — no config file. In ChatGPT, go to Settings → Connectors → Create and paste https://pinescript-mcp.fly.dev/mcp into the Server URL field. Developer Mode must be enabled (see OpenAI Developer Mode guide).
Claude.ai — add via the web UI's MCP connector settings, not a JSON file.
Version Pinning
Documentation is bundled in the package — each version contains a frozen snapshot. For reproducible agent behaviour, pin to a specific version:
{
"mcpServers": {
"pinescript-docs": {
"command": "uvx",
"args": ["pinescript-mcp==0.7.6"]
}
}
}
Without pinning, uvx pinescript-mcp gets the latest version.
Available Tools
| Tool | Description |
|---|---|
resolve_topic(query) | Fast lookup for exact API terms (ta.rsi, repainting) |
search_docs(query) | Grep for exact strings across all docs |
list_docs() | List all documentation files with descriptions |
list_sections(path) | List ## headers in a doc file (for navigating large files) |
get_doc(path) | Read a specific documentation file |
get_section(path, header) | Read a specific section by header |
get_functions(namespace) | List valid functions (ta, strategy, etc.) |
validate_function(name) | Check if a function exists in Pine v6 |
list_resources() | Browse available documentation resources |
read_resource(uri) | Read a doc resource by URI (e.g. docs://manifest) |
list_prompts() | List available prompt templates |
get_prompt(name, arguments) | Render a prompt template with arguments |
Available Prompts
| Prompt | Description |
|---|---|
debug_error(error, code) | Analyze a Pine Script compilation error |
convert_v5_to_v6(code) | Convert Pine Script v5 code to v6 syntax |
explain_function(name) | Explain a Pine Script function in detail |
Available Resources
| URI | Description |
|---|---|
docs://manifest | Start here — routing guide for Pine Script questions |
docs://functions | Complete Pine Script v6 function list (JSON) |
docs://{path} | Any doc file by path (e.g. concepts/timeframes.md) |
Example Queries
- "How do I create a trailing stop in Pine Script?"
- "What's the difference between var and varip?"
- "Is ta.supertrend a valid function?"
- "How do I avoid repainting with request.security?"
Documentation Coverage
The server bundles comprehensive Pine Script v6 documentation:
- Concepts: Execution model, timeframes, colors, methods, objects, common errors
- Reference: Types, variables, constants, keywords, operators, annotations
- Functions: Technical analysis (ta.*), strategies, requests, drawings, collections
- Visuals: Plots, fills, shapes, tables, lines, boxes, backgrounds
- Writing Scripts: Style guide, debugging, optimization, limitations
Why Use This?
AI models often hallucinate Pine Script functions or use deprecated v5 syntax. This MCP server grounds the AI in actual v6 documentation, preventing:
- Made-up function names (e.g.,
ta.hulldoesn't exist, useta.hma) - Deprecated syntax from v4/v5
- Incorrect parameter orders
- Missing required arguments
Skills
Combine with skills for even more control available at bouch.dev/products/pine-strategy-builder.
Development
# Clone and install locally
git clone https://github.com/paulieb89/pinescript-mcp
cd pinescript-mcp
pip install -e .
# Run the server
pinescript-mcp
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.