Phase 1: The Tool Landscape · 10 min · Any new AI coding tool

Evaluating Any New Tool in Ten Minutes

A new tool launches every week. You don't need to learn them all. You need a 10-minute test that tells you if it's worth your time.

Real-world application: What changes when you use this: you can evaluate any new AI coding tool in 10 minutes — which means you stop chasing hype and start adopting tools based on evidence.

What you will learn

Use a repeatable 10-minute checklist to evaluate any new AI coding tool
Test whether a tool can build, handle corrections, and deploy
Decide when to adopt a new tool vs. when to wait
Avoid hype-driven tool adoption

Objective

Your mission

Run a standardized 10-minute evaluation on a new AI coding tool you haven't tried. Make an adopt/wait/skip decision with evidence. Output: tool_evaluation.md.

You need: A new AI coding tool you haven't tried. A 10-minute timer.

Time: 10 minutes (literally).

Step 1: categorize it (minute 1-2)

Ask: which of the four categories does this fit?

Browser app builder? IDE-native? Terminal agent? Design-to-code?
If it doesn't fit any, it's either genuinely novel (rare) or confusingly marketed (common).

You now know its strengths and limitations from the category alone.

Write in your file:

## Tool: [name]
Category: [which one]

Step 2: the build test (minute 3-5)

"Build a single-page tip calculator. Bill amount input, tip percentage buttons (15%, 18%, 20%, 25%), number of people, live per-person totals. Clean, modern, mobile-first. No backend."

This is your benchmark. You've seen what good tools do with this prompt.

Write in your file:

## Build Test
- Generated a working app? [yes/no]
- Got the features right? [yes/no]
- Clean UI? [yes/no]
- Time taken: [estimate]

Step 3: the correction test (minute 6-7)

Give a specific correction:

"Make the tip buttons a horizontal row of pills instead of a vertical list. Highlight the selected one in green. Keep everything else the same."

Write in your file:

## Correction Test
- Made the change correctly? [yes/no]
- Broke anything else? [yes/no]
- Understood without clarification? [yes/no]

Step 4: the deploy test (minute 8-9)

Can you deploy? How easy is it?

One-click deploy to a live URL? ✓
Requires configuration? ⚠
No deploy option? ✗

Write in your file:

## Deploy Test
- Deploy option: [one-click / requires config / none]

Step 5: the verdict (minute 10)

Score it:

Build passed + Correction passed + Deploy easy → Adopt. Add it to your toolkit.
Build passed but correction struggled → Wait. Note it, check back in 3 months.
Build failed or correction broke things → Skip.
No deploy → Limited usefulness. Only worth it if you have another deploy path.

Write in your file:

## Verdict: [Adopt / Wait / Skip]
Reasoning: [reference specific test results, not vague impressions]

The hype filter
Demo videos are curated — the tool's best output, cherry-picked, edited, sped up. Your 10-minute test uses the same standard prompt every time, so you're comparing tools on a level playing field. A tool that can't build a tip calculator and handle one correction isn't going to build your real project, no matter how good the demo looked.

A new tool builds the standard tip calculator well, but your one correction breaks other features. What's the verdict?

Build passed but the correction struggled is a Wait. A tool that can't take one targeted correction won't handle a real project's iteration.

Key Terms & Next

You're done

Save your tool_evaluation.md. You now have a repeatable process for evaluating any new tool in 10 minutes. Use it every time a new tool launches with a viral demo.

That's the whole lesson. Now do the lab.

Its hands-on lab, and the lessons after it, open with a free account — the opening phases of 29 of 35 courses are free, no credit card.