← Back to Skills Marketplace
unisound-llm

unisound-med-teach

by Unisound-LLM · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
43
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install unisound-med-teach
Description
临床教学案例生成。由调用方提供题目文本或含 `question` 字段的结构化输入,经内部医疗大模型按题干约束生成作答;仅含 `scripts/run.py`,可独立拷贝部署。
README (SKILL.md)

临床教学案例生成

概述

基于给定病历要点生成教学案例,突出学习目标、鉴别点与伦理边界(注意脱敏)。

本能力与具体业务请求解耦:平台或调用方如何组织用户输入由集成方决定;此处只约定如何把题目交给脚本(见下)。

题目输入(三选一)

  1. --question:直接传入完整题干字符串。
  2. --input PATH
    • .jsonl:每行一个 JSON 对象,须含非空字符串字段 question;可用 --index / --batch / --match-id 选取记录;
    • .json:单个对象,须含 question
    • 其他扩展名:按 UTF-8 纯文本整文件作为题干。
    • - 时从 stdin 读取一条 JSON(含 question)或纯文本。
  3. stdin(且未传 --question / --input、且 stdin 非 TTY):读入 JSON 对象或纯文本,规则同 -

模型

  • 默认 u1-insuremed,接口 https://maas-api.hivoice.cn/v1/chat/completions(可通过参数覆盖)。

用法示例

# 命令行直接给题
python3 scripts/run.py --question "题干……" --appkey YOUR_KEY

# 从任意路径的 jsonl 取一条(与文件名无关)
python3 scripts/run.py --input /path/to/items.jsonl --index 0 --appkey YOUR_KEY

参数

  • --question / --input:见上,与 stdin 三选一(--question--input 不可同时使用)。
  • --appkey STRING必填--dry-run 时除外)。内部医疗大模型鉴权 key。
  • --index--match-id--batch:仅对 .jsonl 生效。
  • --dry-run:不调用模型,仅输出解析后的题目 JSON。
  • --model--api-url--temperature--timeout--system-prompt:调模型参数。
  • --output PATH:另存完整 JSON(批量为 NDJSON)。
  • --text-only:标准输出仅打印模型答案文本。

输出

默认可解析 JSON:含 statusquestionanswerrecord_indexmetamodelinput_modeinput_path(来自文件时非空)等;skill 为本任务标题。

合规说明

输出为模型辅助信息,不构成正式诊疗决策;涉及真实患者数据须先脱敏并遵守院内流程。

Usage Guidance
Install only if you are comfortable sending the provided case text to the listed medical LLM API. Use de-identified teaching cases, protect the app key, verify the publisher/metadata mismatch is expected, and store any output files securely.
Capability Analysis
Type: OpenClaw Skill Name: unisound-med-teach Version: 1.0.0 The skill bundle is a legitimate tool for generating clinical teaching cases by interfacing with a medical LLM API (hivoice.cn). The Python script `scripts/run.py` uses standard libraries to handle input from command-line arguments, files, or stdin and performs a standard POST request to the specified API. There are no signs of data exfiltration, unauthorized execution, or malicious instructions in the documentation.
Capability Assessment
Purpose & Capability
The documented purpose and visible code align: it takes a caller-provided question/file/stdin, calls a disclosed medical LLM endpoint, and returns an answer. The main user consideration is that clinical case text may be sensitive.
Instruction Scope
The instructions describe normal user-directed inputs and options; they do not try to override agent goals, force hidden tool use, or make untrusted text authoritative.
Install Mechanism
There is no install spec, dependency download, shell setup, or package installation. Provenance is weaker because the source is unknown and embedded _meta.json uses a different owner/slug from the registry metadata.
Credentials
The script reads user-selected input files or stdin, can write a user-selected output file, and sends prompts to the disclosed API endpoint. These capabilities are proportionate to the stated purpose but should be used with de-identified clinical text.
Persistence & Privilege
No background persistence or credential storage is shown. However, the skill requires an app key for non-dry-run use and optional outputs can store the full question, metadata, answer, and input path.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install unisound-med-teach
  3. After installation, invoke the skill by name or use /unisound-med-teach
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of clinical teaching case generation skill. - Generates teaching cases based on key points provided in question text or structured input. - Supports flexible input methods: direct question string, input file (`.json`, `.jsonl`, plain text), or standard input. - Utilizes internal medical LLM (`u1-insuremed`) via configurable API. - Includes options for input selection, model parameters, and dry-run mode. - Outputs structured JSON with answer, input details, and metadata. - Emphasizes de-identification and compliance with patient data protocols.
Metadata
Slug unisound-med-teach
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is unisound-med-teach?

临床教学案例生成。由调用方提供题目文本或含 `question` 字段的结构化输入,经内部医疗大模型按题干约束生成作答;仅含 `scripts/run.py`,可独立拷贝部署。 It is an AI Agent Skill for Claude Code / OpenClaw, with 43 downloads so far.

How do I install unisound-med-teach?

Run "/install unisound-med-teach" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is unisound-med-teach free?

Yes, unisound-med-teach is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does unisound-med-teach support?

unisound-med-teach is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created unisound-med-teach?

It is built and maintained by Unisound-LLM (@unisound-llm); the current version is v1.0.0.

💬 Comments