Skill
unisound-followup-record
将中文门诊复诊病历文本结构化为细粒度字段，输出 JSON（如现病史/既往史/诊断/处理意见等）。
When to use unisound-followup-record
Choose if
You need to convert Chinese outpatient follow-up notes into JSON with 20+ medical fields (present illness, past medical history, marital/menstrual history, physical exam, auxiliary tests, diagnosis, treatment plans). The skill normalizes missing fields to the placeholder "未提及" so downstream parsers see uniform shape. Pick this over a free-form LLM prompt when the consumer expects deterministic field names.
Avoid if
You don't operate in Chinese, you can't provision a Unisound `appkey`, or you need live clinical advice — the README scopes the skill to "text extraction and structuring only; does not constitute medical diagnosis or treatment advice." Also avoid if your records cannot be de-identified before sending to the upstream API.
Risk Flags
- MEDIUM auth Requires an internal `appkey` (Bearer token) for the Unisound maas-api.hivoice.cn endpoint. No public free tier disclosed.
- MEDIUM scope README explicitly states the skill is for text extraction and structuring only, not medical diagnosis or treatment advice.
- LOW scope Chinese-language outpatient records only; missing fields are emitted as the literal string "未提及" rather than null, which downstream JSON consumers must handle.
Cost
Type: Unknown
Distribution
- ClawHub
unisound-followup-record- License
- MIT-0