Learning path

Prompt engineering - foundations that hold up across vendors

Anyone who can already use a chatbot daily and wants to write prompts a teammate can run unchanged.

8 steps92 min total0/8 steps live

By the end of this path

You will (1) write Role-Context-Constraints-Format prompts, (2) know when to add a few-shot example vs a system prompt, (3) maintain a versioned prompt library, and (4) recognise when a reasoning model is worth the cost.

Your progress

Loading your progress…

Walk the path

  1. 1
    L3 · Daily user10 minComing soon
    Prompting fundamentals - specific, context, examples

    Why: The baseline everything else builds on.

  2. 2
    L4 · Crafted user15 minComing soon
    Structured prompting (Role / Context / Constraints / Format)

    Why: The RCCF skeleton - your default template.

  3. 3
    L4 · Crafted user10 minComing soon
    Few-shot examples - when zero-shot is not enough

    Why: When zero-shot fails, examples beat instructions.

  4. 4
    L4 · Crafted user10 minComing soon
    System prompts and persona priming

    Why: Persona priming - the under-used reliability lever.

  5. 5
    L4 · Crafted user10 minComing soon
    Assistant memory and custom instructions

    Why: Set-and-forget context that compounds.

  6. 6
    L4 · Crafted user15 minComing soon
    Build your prompt library

    Why: A library, not a notebook - versioned, reviewed, shared.

  7. 7
    L4 · Crafted user12 minComing soon
    Chain-of-thought reasoning patterns

    Why: Make the model show its working when the answer must be defended.

  8. 8
    L4 · Crafted user10 minComing soon
    Reasoning models - when to use o-series and thinking modes

    Why: Reasoning costs 10x - know when it pays for itself.

Next path
Developer AI fluency - from autocomplete to design partner
Continue

Hi, I'm Nova. Chat, speak, or show me — I'll point you at the right tool.