JuneYaooo/nihaisha-nishi-tcm

倪海厦中医课程资料的 Agent Skill：支持课程检索、方证穴位辨析、学习笔记整理与板书截图证据索引。 | An Agent Skill for Ni Haisha TCM course study, formula-pattern lookup, acupoint reference, and screenshot evidence indexing.

What it solves

This project transforms high-density Traditional Chinese Medicine (TCM) course materials from Ni Haisha into a searchable, traceable, and safe Agent Skill. It solves the problem of navigating vast amounts of course data—including videos, audio, and notes—by allowing users to retrieve specific information using natural language, while ensuring every conclusion is backed by evidence (screenshots or PDF pages).

How it works

The project uses a "course distillation" method to organize TCM materials into a structured format that AI agents (like Claude Code, Codex, or OpenClaw) can use. It maps natural language queries (e.g., "cold with chills") to specific TCM diagnostic terms and retrieves information across multiple modules including the Shanghan Lun, Jingui Yaolue, and acupuncture courses. It provides a multi-layered evidence system:

  • Screenshot Evidence: Nearly 3,000 compressed WebP images of blackboard notes and PPTs.
  • PDF Evidence: Page-level indexing for course scripts and ancient texts.
  • Text Evidence: Independent sections from recommended supplementary materials.

Who it’s for

  • TCM Students: Those studying Ni Haisha's courses who need to review specific modules, compare formulas, or find specific blackboard notes.
  • TCM Researchers: Individuals looking to verify quotes or trace the origins of specific TCM theories in the course materials.
  • AI Agent Users: People who want to integrate a specialized TCM knowledge base into their AI assistants.

Highlights

  • Traceable Evidence: Every answer can be linked back to a specific PDF page or screenshot index.
  • Broad Coverage: Spans multiple core TCM modules including Shanghan Lun, Jingui Yaolue, acupuncture, and Shennong Bencao.
  • Comprehensive Evaluation: Includes a 240-question multi-dimensional test set to measure retrieval accuracy and safety.
  • Safety Boundaries: Explicitly restricts the agent from providing personal medical diagnoses, prescriptions, or dosage guidance.
  • Natural Language Interface: Converts colloquial descriptions of symptoms into professional TCM diagnostic categories.

Related

  • Project
    jangviktor-web/nihaixia

    An AI Agent skill that distills the complete TCM thinking system of Ni Haixia, enabling AI to perform Six Meridian differential diagnosis and formula prescription based on 3.5M words of distilled lectures and cases.

  • Project
    KKKKhazix/khazix-skills

    A collection of standardized AI agent skills (instruction sets) for goal definition, disk cleanup, deep research, and project documentation maintenance.

  • Project
    larashero3-dotcom/lieflat-less-ai-tone

    A linguistic research project and toolset that quantifies "AI flavor" in Chinese text and provides a set of rules to remove these markers for more human-like writing.

  • Project
    GiMi-Xiaomi/gimi-illustration-skill

    An AI Agent Skill that converts text articles and scripts into conceptual illustrations using customizable styles and consistent character IPs.

  • Project
    KKKKhazix/human-writing

    A framework and toolset for AI agents to eliminate "AI-style" Chinese writing by enforcing human-like rhythms, factual substance, and structural checks.