Introduction#

OneTick is a high-performance market data analytics platform available as both a managed cloud service and an on-premises deployment.

The OneTick MCP server exposes OneTick data and functionality to AI assistants via the Model Context Protocol standard. OneTick operates a hosted instance of this server configured for OneTick Cloud, with the full set of reference and search tools enabled. This documentation covers self-hosted deployment — giving customers full control over which tools are enabled, where the server connects (OneTick Cloud or a private OneTick server), and what documentation and reference data it exposes.

You install and run onetick-mcp, a thin wrapper around the underlying engine, onetick-llmbase — see its PyPI page for the full dependency list.

Capabilities#

The OneTick MCP server lets users interact with the OneTick platform in natural language. Ask it to explore available databases and schemas, search API documentation and usage examples, look up financial instruments by name or identifier, or check trading calendars and market hours — all without writing queries or navigating documentation manually.

Tools are grouped into three categories — Documentation search, Reference data, and Query Based — and are registered dynamically based on the documentation sources and data files you configure. See Tools for a full breakdown.

  • Documentation search performs vector similarity search over indexed OneTick documentation — Python API docs, usage examples, cloud coverage, database catalogue, and SQL reference. Each document set is a collection; the server can index and serve any subset of them. Requires a one-time index build before first use. See Documentation Search Tools.

  • Reference data provides structured access to OneTick database metadata: listing available databases, browsing table schemas, and retrieving field descriptions and enumerations. The server ships with reference files for OneTick Cloud; these can be replaced with your own to match a different deployment. See Reference Data Tools.

  • Query Based — queries a live OneTick WebAPI to look up financial instruments by name, ISIN, MIC, security type, and other attributes. It searches OneTick’s symbol universe — the catalogue of tradable instruments that OneTick tracks across global markets. See Symbol Universe Tool.

Customization#

Self-hosting gives you full control over what the MCP server knows and exposes. Both the documentation content and the reference data can be replaced or extended to match your environment:

  • Custom documentation sources — register additional document sets alongside the built-in collections. Each one is indexed and exposed as a new search tool automatically. See Custom Search Tools.

  • Custom reference data — replace the bundled OneTick Cloud reference files with your own to reflect the databases and schemas available in your deployment. See Reference Data Tools.

Getting started#

The Quickstart walks through installing the server, starting the required PostgreSQL database, building the search index, and running the server end-to-end. For production deployment options (workers, timeouts, health checks) see Deployment.