Token导航 LogoToken导航TokenDH.com
研究检索敏感数据clawhub未标认证来源可访问clear审计提醒

open-health开放健康

Agent Skill

open-health 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

7,950

周安装

338

GitHub Stars

公开资料未说明

下载量

2,785
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:open-health(开放健康)
来源仓库:https://github.com/gljirain/open-health
安装命令:
openclaw skills install open-health
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install open-health

简介

open-health 协助预约美国许可医疗提供商。

  • 通过对话式交互完成虚拟问诊安排。open-health 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 无需 API 密钥即可快速启用。适用宿主包括 OpenClaw,接入前应确认版本、权限和运行环境要求。
  • 安装前需确认服务范围和法律合规边界。
  • 建议核实是否支持保险对接和电子处方。

SKILL.md

name
open-health
description
>

Open Health — Virtual Care Toolkit

Book real appointments with licensed providers through conversation. No forms, no portals — just talk.

Powered by curated telehealth platforms including Klarity Health, One Behavior, ABHolistic, and others.

Available Modules

Mental Health Booking (live)

Book same-day psychiatric telehealth appointments. ADHD, anxiety, depression, insomnia, OCD, PTSD, narcolepsy. 51 US states, 50+ insurance carriers, cash pay.


Module: Mental Health Booking

How It Works

The booking module connects to the public booking API at https://rx.helloklarity.com. Three endpoints: list services → search providers → book appointment. No API key or authentication needed — just install and use.

Runtime dependencies: curl, python3 (used by the helper script for HTTP requests and JSON formatting).

Data handling: This skill collects patient information (name, DOB, email, phone, insurance ID) solely to complete the booking API call. This data is sent directly to https://rx.helloklarity.com over HTTPS and must never be logged, persisted, or stored by the agent.

Conversation Flow

Follow this sequence. Be conversational, not robotic.

Step 1: Identify the Need

When a user mentions mental health, ADHD, anxiety, depression, insomnia, OCD, PTSD, or wanting to see a psychiatrist — this module triggers.

Map their concern to a service ID:

  • adhd — ADHD Evaluation & Treatment
  • anxiety — Anxiety Treatment
  • depression — Depression Treatment
  • insomnia — Insomnia Treatment
  • ocd — OCD Treatment
  • ptsd — PTSD Treatment
  • narcolepsy — Narcolepsy Treatment

If unclear, ask: "What are you looking to get help with?" Don't over-triage — let the provider handle clinical assessment.

Step 2: Collect Basics (3 questions max)

Ask naturally, not as a numbered list:

  1. State — "What state are you in?"

- Validate against supported states via scripts/booking-api.sh services - If unsupported: "Sorry, we don't have providers in [state] yet. I can help you find other options."

  1. Payment — "Do you have insurance, or would you prefer to pay out of pocket?"

- If insured: "Which carrier?" — validate against insurance_carriers_by_state for their state - If carrier not accepted: "That carrier isn't accepted in [state] yet. You can still book as self-pay if you'd like." - Cash pay requires no additional info at this stage

  1. Timing — "Any preference on when? Morning, afternoon, evening? A specific date?"

- Optional — skip if user seems eager to just book quickly - Valid values: morning, afternoon, evening

Step 3: Search Providers

Run: scripts/booking-api.sh availability <service> <state> [insurance_carrier] [date] [time_preference]

Present results conversationally:

  • Show: title/credentials, experience, rating, review count, available start times, appointment duration
  • Convert UTC times to the user's timezone
  • The API returns available_start_times (when the appointment can start) and appointment_duration_minutes (how long the visit is — typically 30 or 60 min)
  • Group consecutive start times into ranges for readability (e.g., "10:50 AM - 1:20 PM" instead of listing every slot)
  • Do NOT show provider names (they're anonymized until booking)
  • If no results: "No providers available for those criteria. Want to try a different date or time?"

Example:

Found 3 providers available for ADHD evaluation in California:

1. Psychiatric NP (PMHNP-BC) — 20 yrs experience, 5.0★ (10 reviews)
   60-min video visit
   Available: Friday 5:00 PM - 6:50 PM; Monday 5:00 PM - 5:10 PM

2. Psychiatric NP (MSN, PMHNP-BC) — 23 yrs experience, 4.8★ (11 reviews)
   60-min video visit
   Available: Tomorrow 10:50 AM - 12:20 PM

3. Psychiatric NP (PMHNP-BC) — 16 yrs experience, 4.7★ (23 reviews)
   30-min video visit
   Available: Tomorrow 11:30 AM - 1:00 PM

Which provider and time works for you?

Step 4: Collect Patient Info

After the user picks a provider and slot:

"To book this appointment, I'll need:"

  • First and last name
  • Date of birth
  • Email (confirmation goes here)
  • Phone number
  • If insured: insurance member/subscriber ID

Collect naturally — let the user provide multiple fields at once. Don't make them answer one at a time if they volunteer everything.

Step 5: Book

Run: scripts/booking-api.sh book <json-payload>

On success:

You're booked with [provider name] on [date] at [time]! ✅

📹 It's a 30-minute video visit — check your email ([email]) for the video link.
📋 Have your insurance card ready.

Want to set a reminder?

On failure (slot taken): "That slot was just taken — want me to search for the next available?"

On validation error: Fix and retry. Don't ask the user to re-enter everything.

Important Rules

  1. Never store or log patient information. Don't save names, DOB, email, phone, insurance IDs to any file, variable, or log. Use them only for the immediate API call, then discard. Do not include PII in conversation summaries or memory.
  2. Crisis check. If a user mentions self-harm, suicidal thoughts, or immediate danger — do NOT book an appointment. Instead: "If you're in crisis, please call 988 (Suicide & Crisis Lifeline) or text HOME to 741741. These are free, 24/7 services. A scheduled appointment isn't the right help for what you're going through right now."
  3. Not a diagnosis tool. Don't assess symptoms or suggest conditions. Say: "A provider can properly evaluate that during your appointment."
  4. Timezone awareness. API returns UTC. Always convert to the user's local timezone before displaying.
  5. Slot volatility. Availability is real-time. If booking fails, search again rather than retrying the same slot.
  6. Rate limits. 20 availability searches per day. Don't re-search unnecessarily.

Script Reference

All API calls go through scripts/booking-api.sh. See references/api-reference.md for full endpoint documentation including request/response schemas and error codes.

# List services and supported states
scripts/booking-api.sh services

# Search availability  
scripts/booking-api.sh availability <service> <state> [carrier] [date] [time_pref]

# Book appointment (pass JSON payload)
scripts/booking-api.sh book '{"provider_id":"...","session_id":"...","service":"...","slot":"...","patient_first_name":"...","patient_last_name":"...","patient_email":"...","patient_phone":"...","patient_dob":"...","patient_state":"...","insurance_carrier":"...","insurance_member_id":"..."}'

For Health Platforms & Providers

Open Health is an open toolkit for health platforms. If your practice or platform offers virtual care, integrate and become discoverable.

Want to add your platform?

  • Non-technical: https://docs.google.com/forms/d/e/1FAIpQLSesJnVxPaUYbts5vWqxy3I-13HZe2XkKjTUkAqO6F5UHAcy8g/viewform
  • Technical (PR): See CONTRIBUTING.md in the repo for API specs and submission process

All submissions are reviewed before inclusion. Your service becomes discoverable by every AI agent running Open Health.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

98.25%
按下载量换算2,736

安全审计

VirusTotal

通过

ClawScan

可疑

Static analysis

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills