avangardistic/MQL5-Knowledge-Graph-System ? reverse-engineered prompt

Build me a Python tool for MQL5 projects that scans a codebase and turns it into a knowledge graph for AI assistants.

I want to point it at an MQL5 folder, have it index the files, understand symbols, includes, callers, callees, references, and basic impact, then let me search and inspect that data from a simple command line. It should also be able to give me a compact context summary for a symbol, with evidence for why each relationship exists, so I can paste that into an AI coding assistant instead of dumping in whole files.

Please make it handle messy or unfinished MQL5 code without breaking, keep the results deterministic, and support incremental reindexing when only a few files change. If it makes sense, include a small API and an MCP server so other AI tools can connect to it. Use the current docs online if you need to check anything about MQL5 or MCP.