Lab 1: Timer via Chat

The first prompt cycle

← Back to overview

Mode: Chat Duration: approx. 10-15 minutes Shows: the first prompt cycle

Runs entirely locally: HTML, CSS, vanilla JavaScript — no external dependencies.

Goal

Start with a deliberately simple prompt, compare the result, then get creative. Participants experience chat mode and learn along the way: an open prompt gives the AI a lot of freedom — results vary.

Preparation

  • An empty folder as a project (in Kiro via Open a project).
  • Kiro in chat mode (default).

Part 1 – The first, open prompt

Everyone enters the same, deliberately plain prompt:

Create a timer using HTML, vanilla JavaScript and CSS without external dependencies.

Observe & compare

  • Try the result briefly (open the file in the browser, see the box below).
  • Compare as a group: the results will differ — countdown vs. stopwatch, different layout, different buttons, different colors.
  • Discussion: Why? An open prompt leaves a lot of room for interpretation. The more precise the prompt (format, requirements, examples), the more predictable the result.

How to open the file in the browser: In the Kiro Explorer (left), right-click the HTML file → Reveal in File Explorer, then double-click it in Windows Explorer. It opens in the default browser. After changes, reload the page with F5.

Part 2 – Now get creative

No fixed prompt — participants design their timer to their own idea. They can rework the existing timer (follow-up prompt) or start from scratch.

Examples for inspiration (just ideas, feel free to find your own):

  • „Turn the timer into a Rolex — an analog clock with hands and a classy look.“
  • „Design the timer in Star Trek LCARS style with the typical colors and shapes.“
  • „Build a Pomodoro timer (25 min work / 5 min break) with a progress ring.“
  • „Make a retro flip-clock timer like an old split-flap display.“
  • „Design a countdown in neon/synthwave style with a glow effect.“

Tip for good results: if you have a clear picture in mind, describe it (style, colors, behavior) — this shows how much more precise prompts steer the outcome.

Key takeaways

  • Chat mode is ideal for fast, exploratory building.
  • Same prompt ≠ same result: open prompts produce different outcomes.
  • Precision (style, format, examples) makes results more predictable.
  • Follow-up prompts build on the existing result (iterative working).