description: Scaffold a new hr-* skill directory with SKILL.md, content/, and examples/ templates
New skill
Ask the user for the skill name. It must start with hr- and use kebab-case only (for example hr-retention, hr-interviewing). If the name does not follow this format, prompt again.
Once you have a valid name, derive these values:
<full-name>— the full name the user provided (e.g.hr-retention)<short-name>— the part afterhr-(e.g.retention), used as the filename for the content doc (nohr-prefix)<Title Case>— a human-readable title derived from<short-name>(e.g.Retention)
Then:
- Create the directory
skills/<full-name>/ - Create
skills/<full-name>/SKILL.mdusing the SKILL.md template below — replace every<full-name>,<short-name>,<Title Case>, and<placeholder>with appropriate content based on the skill name the user provided - Create the directory
skills/<full-name>/content/ - Create
skills/<full-name>/content/<short-name>.mdusing the content template below - Create the directory
skills/<full-name>/examples/ - Ask the user for one realistic scenario name to scaffold the example around (e.g. "designing a retention plan for a startup"), derive
<scenario-slug>from it in kebab-case, and createskills/<full-name>/examples/<scenario-slug>.mdusing the example template below - Make sure the content file's "Example Workflow" section links to
../examples/<scenario-slug>.mdand the example file may reference the content file via../content/<short-name>.mdif relevant - Tell the user: "Skill scaffolded with
SKILL.md,content/<short-name>.md, andexamples/<scenario-slug>.md. Now edit the files to add real content, then runbun run syncfollowed bybun run validate."
Directory structure produced
skills/<full-name>/
├── SKILL.md
├── content/
│ └── <short-name>.md
└── examples/
└── <scenario-slug>.md
SKILL.md template
Use this exact template for skills/<full-name>/SKILL.md (canonical source: .github/skill-template.md):
---
name: <full-name>
description: >
A concise description of at least 50 characters that explains what this skill
covers and when to use it. Include HR trigger phrases such as "write a [task]",
"create a [deliverable]", or "analyze [topic]".
metadata:
author: Tuan Duc Tran
version: "1.0.0"
---
# HR <short-name>
Brief one-paragraph overview of what this skill covers and who it helps.
## Supported tasks
- [Task 1]
- [Task 2]
- [Task 3]
- [Task 4]
- [Task 5]
- [Task 6]
- [Task 7]
- [Task 8]
## Key prompts
### [Subtopic 1]
1. "[Prompt for [context]]"
2. "[Prompt for [role] focused on [area]]"
3. "[Prompt using [variable] to [goal]]"
4. "[Prompt that addresses [scenario]]"
### [Subtopic 2]
1. "[Prompt for [context]]"
2. "[Prompt for [role] focused on [area]]"
3. "[Prompt using [variable] to [goal]]"
4. "[Prompt that addresses [scenario]]"
### [Subtopic 3]
1. "[Prompt for [context]]"
2. "[Prompt for [role] focused on [area]]"
3. "[Prompt using [variable] to [goal]]"
4. "[Prompt that addresses [scenario]]"
## Tips
- [Professional best-practice tip 1]
- [Professional best-practice tip 2]
- [Professional best-practice tip 3]
- [Professional best-practice tip 4]
content template
Use this exact template for skills/<full-name>/content/<short-name>.md:
# Understanding <Title Case> for Strategic HR Management
## Overview
> Related example: [<Scenario Title>](../examples/<scenario-slug>.md).
[One paragraph introducing what <short-name> covers, how it has evolved beyond a
narrow legacy definition, and what modern organizations now include under it.]
For HR professionals, understanding <short-name> is not about [doing everything at
once]. It is about developing the ability to [select the right approach], [read the
right signals], and [connect decisions to business outcomes].
## What <Title Case> Actually Means
Strategic <short-name> in modern organizations typically includes:
- [Capability 1]
- [Capability 2]
- [Capability 3]
- [Capability 4]
- [Capability 5]
- [Capability 6]
- [Capability 7]
<Title Case> today is less about [old framing] and more about
**[new framing, bolded]**.
## Key <Title Case> Categories
### [Category 1]
[One sentence describing what this category measures.]
Focuses on:
- [Sub-area 1]
- [Sub-area 2]
- [Sub-area 3]
### [Category 2]
[One sentence describing what this category measures.]
Focuses on:
- [Sub-area 1]
- [Sub-area 2]
- [Sub-area 3]
### [Category 3]
[One sentence describing what this category measures.]
Focuses on:
- [Sub-area 1]
- [Sub-area 2]
- [Sub-area 3]
## Core <Title Case> Concepts HR Should Understand
### Leading Indicators
Leading indicators are signals that predict future risk before it becomes visible
in lagging data. They allow HR to intervene early.
Examples include:
- [Example 1]
- [Example 2]
- [Example 3]
### Lagging Indicators
Lagging indicators measure outcomes that have already occurred. They confirm what
happened but cannot prevent it.
Examples include:
- [Example 1]
- [Example 2]
- [Example 3]
### [Domain-specific concept pair, e.g. "X Versus Y"]
[One paragraph explaining a key distinction unique to this skill area that is
commonly conflated by practitioners.]
### The <Title Case>-to-Decision Connection
A [metric/assessment/plan] without an owner, a target, and a review cadence does
not generate value. Every [data point] should connect to a specific decision or
intervention.
## How <Title Case> Systems Are Built in Practice
Modern <short-name> frameworks typically follow this structure:
1. [Step 1]
2. [Step 2]
3. [Step 3]
4. [Step 4]
5. [Step 5]
6. [Step 6]
7. [Step 7]
8. [Step 8]
This framework is often supported by:
- AI-assisted analysis tools (Claude, ChatGPT, Gemini)
- [Platform category 1]
- [Platform category 2]
- Early warning protocols for [risk type]
## What HR Should Look For in <Title Case>
### Strong Signals of Effective <Title Case>
- [Signal 1]
- [Signal 2]
- [Signal 3]
- [Signal 4]
- [Signal 5]
### Weak Signals of Ineffective <Title Case>
- [Signal 1]
- [Signal 2]
- [Signal 3]
- [Signal 4]
- [Signal 5]
## Common Misunderstandings in <Title Case>
### [Misunderstanding 1 Title]
[One paragraph explaining the misunderstanding and the more accurate view.]
### [Misunderstanding 2 Title]
[One paragraph explaining the misunderstanding and the more accurate view.]
### [Misunderstanding 3 Title]
[One paragraph explaining the misunderstanding and the more accurate view.]
### [Misunderstanding 4 Title]
[One paragraph explaining the misunderstanding and the more accurate view.]
## Modern <Title Case> Reality
HR teams today are expected to:
- [Expectation 1]
- [Expectation 2]
- [Expectation 3]
- [Expectation 4]
- [Expectation 5]
- [Expectation 6]
Effective <short-name> now requires both:
- [analytical/technical capability]
- and [communication/influence skill]
## Key Industry Benchmarks (2025-2026)
| Metric | Benchmark | Source |
|---|---|---|
| [Metric 1] | [Value] | [Source] |
| [Metric 2] | [Value] | [Source] |
| [Metric 3] | [Value] | [Source] |
| [Metric 4] | [Value] | [Source] |
| [Metric 5] | [Value] | [Source] |
## Example Workflow
See the [<Scenario Title>](../examples/<scenario-slug>.md)
for an example of applying this guide to an end-to-end <short-name> process.
## Conclusion
For HR managers and business partners, the key shift is this:
> [One or two sentence quote summarizing the strategic reframe of this skill area.]
Understanding this distinction improves:
- [Outcome 1]
- [Outcome 2]
- [Outcome 3]
- [Outcome 4]
examples template
Use this exact template for skills/<full-name>/examples/<scenario-slug>.md:
# <Scenario Title>
## Context
[Two to four sentences setting up a realistic company situation: company size and
stage, the trigger event or pressure (CEO/CFO/board question, a flagged risk), and
what data or tooling is currently in place versus missing.]
## Step 1 — [Choosing/Identifying the Right Starting Point]
**Sample prompt:**
> "[A realistic question the user would ask the skill, in their own voice]"
**Expected skill response:**
[Guidance broken into 2-3 named sub-areas, each with a short rationale tying back
to the context above.]
> **Avoid:** [One sentence warning against a common overreach or mistake at this step.]
## Step 2 — [Calculating/Mapping Baseline Data]
**Sample prompt:**
> "[A realistic data-specific question]"
**Skill walkthrough:**
[Show a calculation, classification, or structured breakdown using a fenced `txt`
code block where appropriate.]
**Benchmark context:**
[One paragraph comparing the result to an industry benchmark, citing a source.]
> **[Financial/business impact estimate]:** [One paragraph translating the finding
> into a cost or risk figure.]
## Step 3 — [Diagnosing the Root Cause]
**Sample prompt:**
> "[A realistic diagnostic question]"
**Skill response — diagnostic framework:**
[Four numbered dimensions to investigate, each with 2-3 sub-questions.]
> **What to bring to leadership:** [One sentence on how to present findings credibly.]
## Step 4 — [Building the Core Deliverable]
**Sample prompt:**
> "[A realistic request for a framework, dashboard, or plan]"
**Skill response — [deliverable] structure:**
[Tables or structured sections showing the deliverable, broken into 2-3 parts.]
**Implementation notes:**
- [Note 1]
- [Note 2]
- [Note 3]
## Step 5 — [Presenting to Leadership]
**Sample prompt:**
> "[A realistic question about framing a difficult conversation]"
**Skill response — [stakeholder] communication framework:**
[Three labeled parts: confirm the observation, contextualize it, flag the real risk
and the action plan, each with a sample quoted line of dialogue.]
> **Key principle:** [One sentence on what this stakeholder responds to.]
## Step 6 — Using AI to Analyze the Data
**Sample prompt:**
> "[A realistic question about using Claude/ChatGPT on this data]"
**Skill response — AI-assisted analysis workflow:**
**Step 1 — Prepare your data:**
[What fields/columns to include, and any anonymization guidance.]
**Step 2 — Use this prompt structure:**
```txt
[A reusable structured prompt template the user can paste their own data into]
```
**Step 3 — Follow up with action planning:**
```txt
[A reusable follow-up prompt template]
```
**What AI does well here:**
- [Capability 1]
- [Capability 2]
- [Capability 3]
**What AI cannot replace:**
- [Limitation 1]
- [Limitation 2]
- [Limitation 3]
## Step 7 — [Designing a Forward-Looking Process]
**Sample prompt:**
> "[A realistic question about building something repeatable]"
**Skill response — [process] design:**
[Structured guidance, ideally with a table mapping process steps to owners and cadence.]
> **Common finding at [company type]:** [One paragraph naming a frequent root cause
> and the fastest practical fix.]
## Step 8 — Writing the [Recurring Report] for Leadership
**Sample prompt:**
> "[A realistic request for a reusable report template]"
**Generated report template:**
```text
[A full fill-in-the-blank report template with section headers, separated by
horizontal rule characters, matching the style of the skill's recurring
leadership reporting cadence]
```
## Full <Title Case> Workflow Summary
```text
[Step label]
↓
[Step label]
↓
[Step label]
↓
[Step label]
↓
[Step label]
↓
[Step label]
↓
[Step label]
↓
[Step label]
```
### Common <Title Case> Mistakes
| Mistake | How to avoid it |
|---------|----------------|
| [Mistake 1] | [Fix 1] |
| [Mistake 2] | [Fix 2] |
| [Mistake 3] | [Fix 3] |
| [Mistake 4] | [Fix 4] |
| [Mistake 5] | [Fix 5] |
| [Mistake 6] | [Fix 6] |
| [Mistake 7] | [Fix 7] |
Expert Next.js App Router
Developpement
Un skill qui transforme Claude en expert Next.js App Router.
Générateur de README
Developpement
Crée des README.md professionnels et complets pour vos projets.
Rédacteur de Documentation API
Developpement
Génère de la documentation API complète au format OpenAPI/Swagger.