A hands-on prompting course: how to frame a task so ChatGPT, Claude and Gemini get it right the first time. Prompt structure, context, roles, output formats, chains and the usual mistakes — on real work tasks.
What you'll learn
- Build prompts with a structure: role, task, context, format, constraints
- Control output quality with examples, acceptance criteria and self-checks
- Work with long context and files without losing the point
- Chain prompts for repeatable workflows
- Tell when to use ChatGPT, Claude or Gemini
- Save prompts as reusable templates inside the studio
Curriculum
Module 1. How the model thinks
- Tokens, context and temperature in plain words
- Why the same request gives different answers
Module 2. Anatomy of a working prompt
- Role, task, context, format, constraints
- Examples over explanations: few-shot in practice
Module 3. Controlling quality
- Acceptance criteria and model self-checks
- Iterating on the answer instead of rewriting the prompt
Module 4. Chains and templates
- Splitting a big task into steps
- Your own prompt library for recurring work
Module 5. The right model for the job
- ChatGPT, Claude, Gemini: strengths of each
- Practice: one brief, three models, side by side
FAQ
Do I need a technical background?
No. A prompt is plain language. You need to state a task clearly — that is exactly what the course teaches.
Which model should I practise on?
Any of the ones in the Zerocoder studio: ChatGPT, Claude and Gemini sit in one window, one click apart.
Is this a list of "secret prompts"?
No. Prompt lists go stale in a month. This is the underlying method: how to build a prompt for your own task.
How long does it take?
About three hours including practice. Modules are short and self-contained.
