Home

# 🧬 Welcome to Taya Skills

> Open-source AI Agent Skills for Claude — built by Taya Intelligence

[![License](https://img.shields.io/badge/License-MIT-10B981?style=flat-square)](https://github.com/TayaIntelligence/skills/blob/main/LICENSE)
[![Website](https://img.shields.io/badge/🌐_yiya--ai.com-059669?style=flat-square)](https://www.yiya-ai.com)

---

## 🧰 Skills Overview

| Skill | Type | What it does |
|-------|------|---------------|
| `technical-pm` | Generative | Turns raw ideas into structured PRDs with AI/ML Product Lens built in |
| `prd-review` | Review | Audits PRD completeness, outputs an interactive HTML report |
| `technical-pm-guided` | Teaching | Socratic coaching — learn PM skills by doing, with real-time feedback |

**Supported platforms:** Claude Code · Claude.ai · Claude API · Markdown-compatible AI tools (Codex, etc.)

---

## 🚀 Quick Start

```bash
git clone https://github.com/TayaIntelligence/skills.git

In Claude Code

/technical-pm "Design a clinical trial monitoring tool for CRAs"
/prd-review ./path/to/prd.md
/technical-pm-guided

In Claude.ai

Open Claude.ai, create a new conversation, paste the contents of skills/<skill-name>/SKILL.md at the top, then start chatting.

Via API

import anthropic

client = anthropic.Anthropic()
with open("skills/technical-pm/SKILL.md") as f:
    skill = f.read()

client.messages.create(
    model="claude-sonnet-4-5",
    system=skill,
    messages=[{"role": "user", "content": "Help me design a clinical data platform"}]
)

---

📋 Skills in Detail

technical-pm

8 workflows covered:

1. Discovery & framing
2. PRD / specs
3. Decomposition
4. Prioritization
5. Estimation
6. Metrics
7. Risk
8. Communication

AI/ML Product Lens included: eval set design, threshold selection, precision/recall tradeoffs, model iteration strategy.

prd-review

6 review dimensions: Completeness · Clarity · Consistency · Executability · Measurability · Risk assessment

Verdict: 🟢 Ready / 🟡 Conditional / 🔴 Not Ready

Output: Self-contained HTML report you can share with your team.

technical-pm-guided

Socratic-style coaching. Instead of doing the work for you, the agent asks guiding questions to help you produce PM artifacts yourself, then gives feedback on your output.

---

📁 Skill Structure

Each skill follows this convention:

skills/<skill-name>/
├── SKILL.md          # Main skill instructions (Markdown)
├── README.md         # Human-readable overview
├── references/       # Lazy-loaded reference material
└── examples/         # Usage examples

---

🤝 Contributing

We welcome contributions.

1. Fork the repo
2. Create a branch
3. Add or improve a skill (follow the structure above)
4. Open a PR

---

📄 License

[MIT](https://github.com/TayaIntelligence/skills/blob/main/LICENSE)

---

Built by Taya Intelligence — AI Native CRO, part of Tigermed · Hangzhou, China

```