Prompt Engineering

This is the canonical guide for writing better prompts. It covers the techniques that actually matter: few-shot learning with examples, chain-of-thought reasoning, systematic optimization through A/B testing, reusable templates, and system prompt design. The progressive disclosure pattern is especially useful, showing you when to add complexity versus keeping it simple. Use this when your prompts are inconsistent, when you're burning tokens on bloated instructions, or when you need to debug why an agent isn't behaving as expected. The concrete examples and error recovery patterns make it practical rather than theoretical.

skills.sh installs: 604
GitHub stars: 33.2K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add davila7/claude-code-templates --skill prompt-engineering --agent claude-code

Installs into .claude/skills of the current project.

Files

Prompt Engineering Patterns

Advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability.

Core Capabilities

1. Few-Shot Learning

Teach the model by showing examples instead of explaining rules. Include 2-5 input-output pairs that demonstrate the desired behavior. Use when you need consistent formatting, specific reasoning patterns, or handling of edge cases. More examples improve accuracy but consume tokens—balance based on task complexity.

Example:

Extract key information from support tickets:

Input: "My login doesn't work and I keep getting error 403"
Output: {"issue": "authentication", "error_code": "403", "priority": "high"}

Input: "Feature request: add dark mode to settings"
Output: {"issue": "feature_request", "error_code": null, "priority": "low"}

Now process: "Can't upload files larger than 10MB, getting timeout"