Pharmax Skills 是一套可移植的 Agent Skill 与聊天启动器集合,通过访谈制药专业人士来生成可复用、由人治理的工作情境文件,让通用 AI 从明确的职责、来源、约束和人工审核边界开始工作,而不是凭空猜测。

Stars

3

7 天增长

暂无数据

Fork 数

0

开放 Issue

6

开源协议

MIT

最近更新

2026-07-30

AI 仓库情报摘要
FR-AI / ANALYSIS

为什么值得关注

它抓住了任务型制药 AI 技能之前缺失的“情境层”,将专业工作知识转化为带有安全边界和人工审核规则的透明工作包,并附有可复现的基准测试与合成演示。

适合谁使用

  • 临床药理学、药物计量学、生物统计、临床开发、监管事务、医学事务等制药和生物技术专业人士
  • 为生命科学构建可移植 Agent Skill 的 AI/ML 工程师和平台团队
  • 希望获得可复用 AI 工作情境的临床运营和数据科学团队
  • 关注制药 AI 治理与合规的合规和质量审查人员

典型使用场景

  • 通过附加 Word 或 Markdown 启动器并回答简短访谈来创建初始专业工作情境
  • 使用现有情境和允许的项目源来刷新或协调项目胶囊
  • 在应用任务特定制药技能之前导出组合式 AI 工作包
  • 使用仓库自带的基准协议和评分量规评估 AI 情境质量

项目优势

  • 可移植:既可用作 Agent Skill,也可作为 Word/Markdown 启动器用于 ChatGPT、Claude 或 Copilot
  • 人治:强调数据边界、安全内核和人工审核规则
  • 透明:生成变更摘要,并将缺失细节标记为 UNKNOWN
  • 可复现评估:包含基准协议、评分量规和一次注明日期的 Codex 结果

使用前须知

  • 不是临床决策支持、医疗建议、经过验证的 GxP 系统,也不能替代合格的科学与监管判断
  • 宿主兼容性因工具而异;Copilot 的认证网页聊天测试拒绝 DOCX,需改用 Markdown 启动器
  • 基准证据有限:仅基于一次注明日期的 Codex 运行,后端模型标识不可用,不能视为跨模型性能声明

README 快速开始

Pharmax Skills

Portable Agent Skills that teach AI how pharmaceutical professionals work.

Pharmax Skills interviews a pharmaceutical professional and turns their answers into a reusable work-context file. Attach that file before future approved work so a general-purpose AI begins with the professional's role, responsibilities, evidence rules, constraints, and human-review boundaries instead of guessing.

The first flagship skill is pharma-context-builder. It creates and refreshes:

  • a reusable professional work context;
  • a project-specific context capsule;
  • a combined AI working pack; and
  • a transparent change summary.

It works as an open Agent Skill for compatible tools and as a single attachable Word or Markdown starter for ordinary ChatGPT, Claude, or Copilot conversations.

[!IMPORTANT] Use only information permitted in your current AI environment. Never upload patient-level data, credentials, sponsor-confidential material, regulatory submission drafts, or employer-proprietary content to an unapproved service.

Start in 60 seconds

Option 1: ordinary chat

  1. Download Pharmax-Skills-Start.docx.
  2. Attach it to a new AI conversation.
  3. Say: "Start my pharma work context. Interview me."
  4. Answer only what you are comfortable sharing.
  5. Save the generated My-Pharma-Work-Context file and attach it before future work.

The Markdown version is Pharmax-Skills-Start.md. GitHub Copilot's authenticated web-chat test rejected DOCX but accepted the Markdown starter; see docs/COMPATIBILITY.md.

Option 2: project or workspace

Attach your existing work-context file plus permitted project sources and say:

Update my context, create a project capsule, and export one AI working pack. Show conflicts and assumptions before finalizing.

Option 3: Agent Skill

Install or copy the pharma-context-builder directory into a skills location supported by your AI host, then ask:

Use pharma-context-builder to create or update my pharma work context.

Host installation paths differ. Review [`

项目描述

Portable Agent Skills that teach AI how pharmaceutical professionals work.

相关仓库与替代方案

根据分类、Topic 和编程语言匹配的相似项目。

yetone
精选
yetone GitHub avatar

kill-ai-slop

Kill AI Slop is a multilingual field guide and agent skill that catalogues 33 common AI-generated design tells with interactive before-after demos, and provides a tool to scan and fix web projects.

AI 与机器学习AI 智能体
805
gnipbao
精选
gnipbao GitHub avatar

story-to-handdrawn-video

A Remotion-based tool that converts Chinese story text or ordered hand-drawn images into vertical hand-drawn diary-comic animation with handwritten captions, left-to-right reveals, optional page-curl transitions, and silent H.264 output for post-production dubbing.

AI 与机器学习AI 智能体
683
lopopolo
精选
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI 与机器学习AI 智能体
2,390