AI Prompt Engineering Platform

Prompt A/B Playground

Compare two prompts side-by-side and see scored responses in real time

promptlab.codelab.sh/playground

Prompt A Variant A

"Explain recursion to me."

AI Response:

Recursion is when a function calls itself. For example, to calculate factorial(n), you call factorial(n-1). The base case stops the recursion when n reaches 0 or 1…

Clarity

5.5

Depth

4.0

Prompt B Variant B

"You are a CS professor. Explain recursion to a first-year student using a real-world analogy, then show a Python example with comments."

AI Response:

Think of recursion like Russian nesting dolls 🪆. Each doll contains a smaller version of itself until you reach the tiniest one — the base case. Here's how that looks in Python: def fact(n): ...

Clarity

8.8

Depth

8.5

Community Prompt Library

Browse, fork, and rate high-performing prompts from the class

promptlab.codelab.sh/library

Role + Format + Example 9.2 / 10

"Act as a senior engineer. Explain [X] with: 1. Analogy 2. Code 3. Common mistake"

structurerole promptingfew-shot

Chain of Thought 8.8 / 10

"Think step by step before answering. First list assumptions, then reason through each…"

CoTreasoningstructured

Audience Targeting 8.4 / 10

"Explain this concept as if I'm a 10-year-old who loves Minecraft, then as if I'm a PhD…"

audienceclarity

Constraint-Based 8.1 / 10

"Explain X in exactly 3 bullet points. Each must start with a verb. No jargon."

constraintsconcise

Showing 4 of 48 prompts · Load more

AI Rubric Evaluation

Structured criteria scoring with actionable improvement suggestions

promptlab.codelab.sh/rubric

Evaluation Criteria

Specificity

Does the prompt define audience, format, and constraints?

★★★★★

Context

Is background / role information provided?

★★★★★

Output Format

Is the desired response format specified?

★★★★★

Clarity

Is the prompt unambiguous and direct?

★★★★★

8.0

Overall Score / 10

Good — One area for improvement

AI Feedback

✅ Strong output format: Asking for bullet points + code example is excellent — it forces the AI to organize its thinking.

✅ Role context works: "Senior developer" frames the tone and vocabulary well for the audience.

⚠️ Add audience context: Specifying who will read this (e.g., "for a junior developer who knows Python but not CS theory") would significantly improve relevance.

💡 Suggestion: Try adding "…and explain why each algorithm has a different Big-O complexity" to increase depth score.