MCP & Integrations

Model Context Protocol

MCP is the open standard that lets AI assistants use real tools instead of just chatting. I build MCP servers that connect AI to domain-specific data and real functionality.

What is MCP

Model Context Protocol is a standard for connecting AI models to external tools and data. Instead of guessing or hallucinating, the AI calls real APIs, queries real databases, uses real documentation.

Think of it as giving AI hands. The model decides what to do; the MCP server does the doing, with real data and real results.

Model AI Assistant
Protocol MCP
Server Your Tools
Data Real APIs

MCP Servers Built

Property Shared MCP

Live

Full UK property data access. Land Registry transactions, EPC certificates, Rightmove listings, and composite reports — all available to any MCP-compatible AI.

PropertyDataFastAPI
View project →

Property Assistant MCP

Live

Higher-level property assistant. Search, compare, and report on UK properties through natural language via MCP.

PropertyAI Assistant
View project →

PineScript MCP

Live

9 MCP tools for Pine Script v6 — doc search, function validation, AST-based linting, and version migration. 3 built-in prompts (debug, v5→v6 convert, explain). Monitored via Grafana (request volume, latency, availability). Active users via HTTP endpoint.

TradingPineScript
View project →

iRacing MCP App

In Development

Interactive MCP application for iRacing motorsport data. Race results, member stats, reference data, with a UI layer built on the MCP Apps SDK.

MotorsportTypeScriptMCP Apps

Chess MCP App

In Development

Chess teaching system built as an MCP application. Interactive UI with game analysis and educational features.

EducationTypeScriptMCP Apps

MT5 Trading MCP

In Development

15-tool FastMCP server for MetaTrader 5. Bidirectional control — account info, market data, technical analysis, and order execution through AI.

TradingMT5FastMCP

PyP6Xer MCP

Live

MCP server for Primavera P6 XER files. Parse and analyse construction project schedules through AI assistants.

ConstructionP6FastMCP

Context7 Bridge

Live

Bridge connecting Context7 documentation to AI tools. Live, versioned docs access for any library via MCP.

DocumentationPython

Document Processing MCP

In Development

MCP server for document conversion and extraction via the Docling library. PDF, DOCX, and other formats to structured data.

DocumentsPythonDocling

AI Platforms

PydanticAI Platform

Live

Multi-tenant AI agent platform built with PydanticAI. Serve multiple agents to different tenants with typed tool calls and structured outputs.

PydanticAIMulti-tenantFastAPI

Compress PDF/IMG

Live

PDF and image compression tool. Upload, compress, download — built with Python, deployed on fly.io.

PythonUtilityDeployed

Frameworks & Approach

FastMCP

Python framework for building MCP servers. Clean, typed, composable.

PydanticAI

Typed LLM client with structured outputs. Pairs with MCP for reliable tool use.

CrewAI

Multi-agent orchestration. Used for content pipelines and research automation.

MCP Apps SDK

Interactive UI layer for MCP tools. Used in the iRacing and Chess applications.

Build with MCP

If you have domain expertise and want AI tools that actually work with your data, MCP is the way to do it.

Let's build something →

Free Your Main Thread