nautilus-trader MCP server

What this is

An MCP (Model Context Protocol) server that gives an LLM (Claude Code, Codex CLI,etc.) two ways to look things up in nautilus_traderwhile writing strategy code against it:

Code search (index.py) -- structural search over the actual source(Python + Rust), by symbol name. Not fuzzy-text/embedding search: it finds thereal class/function/struct and returns its docstring or full source.
Doc search (rags/) -- semantic (embedding-based) search over theproject's markdown docs (concepts, guides, tutorials), for "how do I..."questions that don't map to a single symbol name.

How it works

index.py spawns two real language servers as subprocesses and talks LSP(JSON-RPC over stdio) to them directly -- pylsp for the Python source undernautilus_trader/python/nautilus_trader, rust-analyzer for the Rust sourceunder nautilus_trader/crates. documentSymbol finds top-levelclasses/functions/structs/impls; hover gets the docstring (Python only --see Known limitations). Results are cached to .code_index_cache.json sincebuilding it costs ~25s (mostly rust-analyzer over ~2600 files).
rags/rag_build.py chunks every file under docs/ and builds adspy.retrievers.Embeddings index (Gemini embeddings), saved to rags/(config.json + corpus_embeddings.npy) plus a shape.json that records theoriginal folder structure so rags/search.py can filter to a subfolder.
mcp_server.py wires both into 5 MCP tools: search_code, get_code_doc,get_code_source, search_docs, show_doc_keys.

Setup

git clone https://github.com/GwangPyo/NautilusTraderMCP.git
cd NautilusTraderMCP
cp .env.example .env   # fill in GEMINI_API_KEY (and OPENAI/ANTHROPIC if you use load_model)
./install.sh           # conda env "mcp" + deps, nautilus_trader clone, code index, doc index

install.sh is idempotent: re-running it skips the nautilus_trader clone, thedocs/ copy, and the (paid) doc-embedding build if they already exist. SetENV_NAME=<name> to use a different conda env name (used for testing so itdoesn't touch the real mcp env).

An env manager isn't required -- uv venv && uv pip install -e . works too;install.sh just standardizes on conda for a reproducible one-command setup.

Register with a client

./add_claude.sh   # claude mcp add
./add_codex.sh    # codex mcp add

Both just point the client at <conda mcp env>/bin/python3 mcp_server.py over stdio.

Known limitations / TODO

.code_index_cache.json and rags/{config.json,corpus_embeddings.npy,shape.json}have no invalidation -- if nautilus_trader/ or docs/ change, you haveto delete the cache files by hand and re-run to pick it up.
No automated tests -- everything so far has been verified by hand(fresh conda env, fresh uv env, real MCP client over stdio).
nautilus_trader/ is cloned from main (unpinned) -- can drift overtime; nothing currently checks it against a known-good commit/tag.

nautilus-trader MCP server