[agenthaulk]

Summary

Promotes the illustrated Chinese beginner guide to a tracked doc and links it from the README.

docs/getting_started_zh.md is an end-to-end first-run walkthrough for a newcomer with zero background: what the tool does, 7 core terms, install (Docker / venv), first launch + control token, a UI tour, mining a first factor, and — importantly — a "what actually happens after you click 开始研究" section (parse → confirm → compute → report, whether the LLM was called, where data comes from, how metrics and the backtest read). It ships 7 flow diagrams and an English/Chinese metric glossary.

This PR:

• adds the guide as a tracked file under docs/ (previously an untracked local draft);
• updates its tab reference to the renamed 已注册因子 label (matching feat(web): rename the 注册表 tab to 已注册因子 #23);
• includes the "switching the LLM provider from the Web UI" section (matching feat(web): runtime LLM provider switching + API key intake from the frontend #22);
• drops it from .dockerignore so it ships in the image alongside the other docs/;
• links it from the README docs list.

Desensitization

Scanned before tracking — no API keys, no control tokens, no absolute local paths, and no internal-attribution or platform-account material. release_safety_scan.py passes on 278 public files (the guide included) and tests/test_release_cleanliness.py is green.