← All missions
MISSION 08 / AI & technology

Automate a small thing

Good automation starts with a process you can explain without a computer.

Your challenge

Build a small spreadsheet, block-code or text-code tool that supports your project. Test normal inputs, edge cases and recovery from an error.

LEARN & APPLY / 1

Describe the rule

Inputs are the information a system receives. Rules transform them into outputs. Start with a written example so another learner can predict the result.

Try it yourself

Write an input-rule-output diagram.

LEARN & APPLY / 2

Build the smallest useful version

Choose a spreadsheet or coding tool that matches the task. One clear working rule is more valuable than many unexplained features. Keep versions so you can compare changes.

Try it yourself

Build a minimal version and save an example input and output.

LEARN & APPLY / 3

Test the awkward cases

Check empty input, unexpected values and boundaries. Explain errors clearly and preserve the user’s work where possible. Do not hide failure behind a success message.

Try it yourself

Record at least five tests, including two edge cases.

WHAT YOU WILL BE ABLE TO DO
  • Clean a small dataset without silently changing its meaning.
  • Explain a rule and test missing, normal and boundary inputs.
WORKED EXAMPLE

See the thinking.

Readings are 10, 12, 11, 50 and blank. The mean of four observed values is 20.75; the median is 11.5. Blank is missing, not zero. The 50 may be genuine or an error: investigate and document your decision rather than deleting it.

RETRIEVE. EXPLAIN. CHECK.

What happens if the blank is silently changed to zero?

Write your own answer before opening the explanation. Then improve one sentence using what you learned.

Compare your reasoning

The mean becomes 16.6 across five values, changing the claim. A missing-value policy and data dictionary must say what is absent, which units apply and how the summary is calculated.

Stretch yourself

Build a spreadsheet or block-code version and compare it with hand calculations. Include five tests and a clear error message.

Build it. Try it. Change one thing.What would you test when the easy example works?Illustrated learning characters · fictional scene
TEN-MINUTE FIRST TRY

Try before you read the solution.

The Builder converts blank input with Number(''), which returns 0. Why can a calculator show a misleading successful result? Write the first validation check.

Scratch space: not saved. Copy useful reasoning into your private workbook.

Short lessons for this project

One project, growing across missions

Choose a fictional club service, a water-data investigation, or an accessible information tool as your capstone. Keep the same problem brief through your budget, prototype, tests, revision and final defence. The crew exercises teach a method; apply that method to your chosen project and explain any adaptation. Simulation is a complete route.

THE CREW AT WORK / GUIDED PRACTICAL

Build a rule you can trace

60–90 minutes · Paper or browser · No purchase required

Builder

“My cost calculator works for 30 units.”

Inventor

“What does it do with zero, a blank entry, −2, or the word ‘thirty’?”

Illustrated learning characters. Scenarios and numbers are fictional; your human mentor reviews your evidence.
01 / PREDICT

Why should a blank input not quietly become zero?

Write a prediction before you reveal the worked reasoning.

Your starter material

Build a cost calculator on paper, in a spreadsheet or JavaScript. Inputs: quantity, price and unit cost. Formula: quantity × (price − unit cost). Use the budget lab as a reference result, not as a substitute for your own build.

PRACTICAL STEP 1

Trace variables and a condition

A variable names a value. Write quantity=3, price=250, unitCost=170. Trace the multiplication to 240. A condition chooses a branch: if quantity is missing, nonnumeric, negative or not whole, show an explanatory error.

PRACTICAL STEP 2

Build and repeat

Implement the formula and validation in your chosen medium. A function groups that operation; a loop can run the same function on several test rows. Do not use eval or execute user-entered code.

PRACTICAL STEP 3

Test boundaries

Record expected and actual results for 3, 0, blank, −2, ‘thirty’, and 1.5 units. Zero is valid; blank and the last four invalid cases must not become a believable total. Fix and rerun failures.

INTERACTIVE / NO ACCOUNT OR AI KEY NEEDED

Cost calculator

Predict a result first. Change an input, compare the output and explain why it changed. These fictional models are learning aids.

Contribution/unit: KES 80.00
Revenue: KES 7500.00
Total cost: KES 8100.00
Surplus / loss: KES -600.00
Break-even: 38 units (check contribution and assumptions)

Tool inputs reset when you leave. Download results or record them in your private mission workbook.

Compare with the crew’s reasoning

For 3 units at 250−170, the result is 240. For 0 it is 0. Number('') produces 0 in JavaScript, so check an empty string before conversion. A good failure message says which input needs correction. Testing one ordinary example misses that bug.

Finding the answer is not evidence of mastery. Explain it, test a changed input and record your own work.

Your evidence checklist

  • I can explain each variable and condition.
  • My log contains normal, zero, blank, negative, text and fractional-quantity cases.
  • Invalid entries produce a useful message and the corrected tests pass.

Independent challenge

Extend to two products using a list and a loop. Explain whether prices may contain decimals while quantities must remain whole.

Open the workbook template
Input | Expected output | Actual output | Pass/fail | Fix
Variables and units:
Rule or pseudocode:
Test 1 normal:
Test 2 zero:
Test 3 blank:
Test 4 negative:
Test 5 text:
Test 6 fractional quantity:
Download the complete offline pack →
A DIFFERENT WAY TO TRY IT

Make the idea your own.

Choose a twist. Use paper, a sketch or a conversation.

Optional practice. Copy useful observations into your workbook; choosing a twist does not record completion.

Compare ideas with someone you know. Sharing is optional.

Need a hint or a clearer next step?

Work through Build a rule you can trace with a focused prompt. Your notes stay private unless you choose to paste an excerpt here.

Use fictional examples. Leave out names, contact details, passwords and private records.

Get an answer here with my own OpenAI API key

Requires an active learning profile. Your provider may charge you; API billing is separate from ChatGPT. EaseBrain has no paid-key fallback.

Key clears after each attempt. EaseBrain saves only a daily request count, not your key, prompt or answer. OpenAI’s data policies apply. Up to 500 output tokens and 20 attempts a day; no automatic retries. These limits do not guarantee a spending ceiling.

Sign in · Complete my learning profile

Ready for a deeper challenge? Request advanced mentorship.

Next mission: Sense. Test. Respond. →