home / skills / nickcrew / claude-cortex / prompt-engineering

prompt-engineering skill

safe

This skill helps you craft, test, and iterate prompts for reliable LLM outputs using structured design and synthetic data.

npx playbooks add skill nickcrew/claude-cortex --skill prompt-engineering

Review the files below or copy the command above to add this skill to your agents.

Files (3)

SKILL.md

2.7 KB

---
name: prompt-engineering
description: Optimize prompts for LLMs and AI systems with structured techniques, evaluation patterns, and synthetic test data generation. Use when building AI features, improving agent performance, or crafting system prompts.
keywords:
  - prompt engineering
  - LLM
  - system prompt
  - few-shot
  - chain-of-thought
  - synthetic data
triggers:
  - prompt engineering
  - system prompt
  - LLM optimization
  - prompt template
  - synthetic test data
---

# Prompt Engineering

Craft, test, and iterate prompts that deliver reliable outputs across LLMs. Covers prompt optimization techniques, structured prompt design, synthetic test data generation, and evaluation methodology.

## When to Use This Skill

- Building or optimizing prompts for AI-powered features
- Crafting system prompts for agents or assistants
- Improving reliability and consistency of LLM outputs
- Generating synthetic test data to validate prompt behavior
- Evaluating prompt performance across edge cases
- Designing prompt chains and pipelines

## Quick Reference

| Task | Load reference |
| --- | --- |
| Prompt techniques and patterns | `skills/prompt-engineering/references/techniques.md` |
| Synthetic test data generation | `skills/prompt-engineering/references/synthetic-data.md` |

## Workflow

1. **Research**: Gather the use case, constraints, and evaluation criteria. Audit existing prompts and model behaviors.
2. **Design**: Draft structured prompts with examples, constraints, and evaluation hooks. Plan experiments and measurement strategy.
3. **Generate test data**: Analyze prompt variables, generate diverse and realistic test cases to validate the prompt.
4. **Validate**: Run prompt trials, capture outputs, document adjustments. Iterate until quality thresholds are met.
5. **Deliver**: Hand off the final prompt with usage guidance and evaluation results.

## Core Principle

When creating prompts, always display the complete prompt text in a clearly marked section. Never describe a prompt without showing it. The prompt must be copyable and self-contained.

## Deliverables Checklist

For every prompt engineering task, produce:

- [ ] The complete prompt text (displayed in full, properly formatted)
- [ ] Explanation of design choices and techniques used
- [ ] Usage guidelines (model, temperature, parameters)
- [ ] Example expected outputs
- [ ] Test cases covering happy path, edge cases, and adversarial inputs

## Example Interactions

- "Optimize this system prompt for our code review agent"
- "Create a prompt for extracting structured data from support tickets"
- "Generate test cases to validate this classification prompt"
- "Design a prompt chain for multi-step document analysis"
- "Improve consistency of this summarization prompt"

Overview

This skill optimizes prompts for large language models and AI agents using structured techniques, evaluation patterns, and synthetic test data. It helps teams design reliable, reproducible prompts and prompt chains that perform consistently across models and edge cases. Use it to improve agent behavior, create system prompts, or validate prompt-driven features.

How this skill works

The skill inspects prompt structure, identifies variability sources, and applies proven patterns like context framing, few-shot examples, and constraint enforcement. It generates synthetic test cases that exercise happy paths, edge cases, and adversarial inputs, then runs evaluation cycles to measure consistency, accuracy, and robustness. Outputs include a final, copyable prompt, design rationale, usage parameters, expected outputs, and a test-suite.

When to use it

Building or optimizing prompts for product features or workflows
Crafting system prompts for agents, assistants, or orchestrators
Improving response reliability and reducing hallucinations
Generating synthetic test data to validate prompt behavior
Designing multi-step prompt chains and pipelines

Best practices

Always publish the complete, copyable prompt text so others can reproduce results
Include few-shot examples that reflect real user inputs and edge cases
Define clear evaluation metrics and acceptance thresholds before experiments
Generate diverse synthetic test cases covering happy, edge, and adversarial scenarios
Document model settings (temperature, max tokens, top-p) and any post-processing steps

Example use cases

Optimize a code-review system prompt to produce consistent, actionable feedback
Design a prompt chain for multi-step document analysis with intermediate QA checks
Create a classification prompt and synthesize labeled test data to validate precision and recall
Craft a system prompt for a customer-support agent and test behavior across adversarial queries
Iterate a summarization prompt to improve factuality and reduce hallucinations

FAQ

What deliverables will I get from a prompt optimization engagement?

A copyable final prompt, explanation of design choices, recommended model parameters, example expected outputs, and a test suite covering happy, edge, and adversarial cases.

Which model settings should I include with the prompt?

Include temperature, max tokens, top-p, any stopping criteria, and suggested decoding strategy; document these to ensure reproducibility.

How do you validate prompt improvements?

Run controlled A/B trials with your test-suite, measure defined metrics (accuracy, consistency, hallucination rate), and iterate until thresholds are met.