← Back to Skills Marketplace
unisound-llm

unisound-med-treat

by Unisound-LLM · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
44
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install unisound-med-treat
Description
专病诊疗方案制定。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)

专病诊疗方案制定

概述

结合病历摘要与约束条件,给出诊断要点、检查与治疗原则的规范化建议框架。

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
Before installing, confirm that you are allowed to send the selected medical case text to the Hivoice/Unisound model API, de-identify patient data, protect the app key, and verify the package provenance because the embedded metadata does not exactly match the registry entry.
Capability Analysis
Type: OpenClaw Skill Name: unisound-med-treat Version: 1.0.0 The skill is a legitimate CLI utility for interacting with the Unisound (hivoice.cn) medical LLM to generate treatment plans. The code in `scripts/run.py` follows standard practices for handling file I/O and making authenticated API requests via `urllib`. There is no evidence of data exfiltration, malicious execution, or prompt injection; it strictly processes user-provided medical questions and returns model responses as described in `SKILL.md`.
Capability Assessment
Purpose & Capability
The stated purpose and visible code align: the skill reads a provided medical question, calls a medical LLM endpoint, and returns JSON/text. Users should still treat medical case text as sensitive and not as a formal diagnosis.
Instruction Scope
Inputs, batching, API URL, model, output path, and dry-run behavior are documented. The configurable API URL and system prompt should remain under user/operator control.
Install Mechanism
There is no install script or package install flow. However, the source/homepage are unknown and the embedded _meta.json owner/slug differ from the registry metadata, so provenance should be verified.
Credentials
The script can read a user-specified file/stdin and write a user-specified output file, then send the selected question to the configured API. This is proportionate for the stated purpose when the user chooses the input deliberately.
Persistence & Privilege
No background persistence or privilege escalation is shown. The skill uses a caller-provided app key and can persist full JSON results only when --output is supplied.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install unisound-med-treat
  3. After installation, invoke the skill by name or use /unisound-med-treat
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of med-eval-med-treat skill for generating standardized diagnosis and treatment frameworks for specific diseases. - Supports flexible input via CLI: direct string, file (json, jsonl, or text), or stdin, with multiple selection mechanisms. - Integrates with internal medical LLM; requires appkey for authentication. - Configurable parameters for model, API endpoint, output format, and more. - Output includes structured JSON with answer, input metadata, and skill identification. - Emphasizes data compliance and non-decisional usage for model outputs.
Metadata
Slug unisound-med-treat
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is unisound-med-treat?

专病诊疗方案制定。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 44 downloads so far.

How do I install unisound-med-treat?

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

Is unisound-med-treat free?

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

Which platforms does unisound-med-treat support?

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

Who created unisound-med-treat?

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

💬 Comments