Home

🧬 Welcome to Taya Skills

Reusable Agent Skills for Claude — built by Taya Intelligence

Each skill is a self-contained folder with a SKILL.md instruction file, plus templates and references it loads on demand.

---

🧰 Skills

technical-pm — Technical Product Manager

Makes the assistant operate like an experienced PM. Give it a fuzzy idea or meeting notes, and it produces clear, scoped, buildable product artifacts. AI/ML product lens built in.

8 workflows: Discovery → PRD → Slicing → Prioritization → Estimation → Metrics → Risk → Communication

Outputs: PRD · epics & user stories · prioritized backlog · story points & sprint plan · metrics · risk register · stakeholder updates

prd-review — PRD Completeness Review

Audits an existing PRD. Answers one question: can an engineer build the right thing from this without guessing?

Verdict: 🟢 Ready / 🟡 Conditional / 🔴 Not Ready

Output: Scored assessment + prioritized gap list + HTML report

technical-pm-guided — Guided Learning

A Socratic coach. Instead of producing artifacts for you, it teaches you to produce them yourself — one step at a time, with feedback.

	technical-pm	technical-pm-guided
Role	PM practitioner	PM coach
Trigger	"Write a PRD"	"Teach me how to write a PRD"
Output	Finished artifacts	You produce, coach gives feedback

---

🚀 Quick Start

Claude Code

cp -R technical-pm ~/.claude/skills/technical-pm
cp -R prd-review ~/.claude/skills/prd-review
cp -R technical-pm-guided ~/.claude/skills/technical-pm-guided

Then invoke:

/technical-pm
/prd-review
/technical-pm-guided

Other AI Tools

Load a skill's SKILL.md as the system instruction. The skill pulls from references/ and assets/ on demand.

---

📁 Repository Structure

skills/
├── technical-pm/
│   ├── SKILL.md
│   ├── agents/openai.yaml
│   ├── assets/          # prd-template.md, risk-register-template.md, user-story-template.md
│   └── references/       # ai-ml, discovery, metrics, prioritization, worked-example...
├── prd-review/
│   ├── SKILL.md
│   ├── agents/openai.yaml
│   ├── assets/          # + report-template.html
│   └── references/       # + review-rubric.md, worked-review-example.md
└── technical-pm-guided/
    ├── SKILL.md
    ├── agents/openai.yaml
    ├── assets/          # shared with technical-pm
    └── references/       # shared with technical-pm

Conventions: SKILL.md is the sole entry point. references/ loads on demand. assets/ holds templates. No extra READMEs inside skill folders.

---

📄 License

MIT