Summary  
This chapter covers prompt engineering techniques: providing detailed context, specifying constraints, and including example templates to guide an AI model’s output more precisely.  

General domain of usage  
Email autoresponders.

Better prompts lead to better results. Be specific, add constraints, and include examples.

### Start specific, then get more specific

The unspecific version of the prompt looks like this:

- **Sarah needs an out-of-office email for a two-week leave.**

Claude can write a professional response, but without enough context, it may use placeholders. For better results include Sarah's role, company, leave dates, backup contact and any special instructions. You can also request a tone such as **professional**, **warm**, or **brief**.

### Adding constraints

Constraints are what stop Claude from over-writing. They define the length, format, and things you want to avoid. For example:

- **Warm but not chummy. No 'thanks in advance'. No 'I hope this finds you well'. Just one short paragraph for the cover details and one closing line.**

Now the result is much more focused and concise.

> Task
>
> - Take a task you've recently used AI for where the result was only okay. Rewrite the prompt with three additions:
> 
>   * A sentence of context about who you are and the situation;
>   * One specific constraint (length, tone, or format);
>   * One example of the style you want.

Learn how to get the most out of Claude for work, study, and everyday tasks. This course covers everything from writing better prompts and analysing documents to using advanced features such as Projects, Artifacts, Memory, Skills, and Connectors. You'll also explore how Claude integrates with tools like Chrome, Slack, Excel, PowerPoint, and mobile devices. By the end of the course, you'll be able to use Claude confidently to save time, improve productivity, and solve real-world problems more effectively.

Learn the fundamentals of Claude, including account setup, the interface, plans, and core capabilities.