### Using with Your AI Assistant

Just tell your assistant:

> "I need to prepare for a 1-on-1 with an underperformer. Use the leadership-prompts skill to find the right prompt, then walk me through it."

Or for browsing:

> "Show me all the hiring prompts from leadership-prompts"

## How to Use a Prompt

Each prompt has **placeholder variables** in `{curly_braces}`. Fill these in with your specific context. The more specific you are, the better the output.

**Example:**

The prompt says:
> I'm preparing for a 1-on-1 with a direct report who has been underperforming for the past {timeframe}...

You fill in:
> I'm preparing for a 1-on-1 with a direct report who has been underperforming for the past 6 weeks...

## Prompt Design Principles

These prompts are designed to:

1. **Force structure** — They use frameworks (SBI, RACI, RAG status) so your output is actionable, not rambling
2. **Include the uncomfortable parts** — Like "prepare for their defensive reaction" or "know when to escalate to HR"
3. **Be opinionated** — "The exec update is your team's marketing" isn't neutral advice, it's a perspective earned from experience
4. **Output something usable** — Every prompt specifies an output format you can immediately use in a meeting, doc, or conversation
5. **Acknowledge politics** — Real leadership happens in political contexts. These prompts include stakeholder dynamics, not just best practices

## Adding Your Own Prompts

Add entries to `prompts.json` following the existing schema:

# List all categories
node scripts/leadership-prompts.js list

# Get a random prompt (great for manager skill-building)
node scripts/leadership-prompts.js random

# Search by keyword
node scripts/leadership-prompts.js search "promotion"

# Show a specific prompt by ID
node scripts/leadership-prompts.js show career-dev-promotion

# Get all prompts in a category
node scripts/leadership-prompts.js category "Team Health"

{
  "id": "category-short-name",
  "category": "Category Name",
  "title": "Human-readable title",
  "prompt": "The actual prompt text with {variables}",
  "context": "When to use this prompt",
  "output_format": "What the AI should produce",
  "example": "A filled-in example showing real usage"
}