Free AI learning · Mission 1 of 3
Give useful context
Choose a task brief that keeps the information needed for a useful draft and leaves unrelated private details out.
Original fictional records and scripted feedback. No model is running; no message, code or approval is sent.
- 1First case
- 2Changed case
- 3Takeaway and quiz
Building software · First case · guided practice
A bounded bug report
Choose a brief for a bug-report draft using only the supplied synthetic fixture.
Fictional issue notes · D1–D3
D1 · Reproduction
In the demo basket, setting quantity to 0 leaves the previous total visible. Expected total: 0 credits.
D2 · Scope
Draft a reproduction report and a proposed test. Do not edit production or claim the cause is known.
D3 · Unrelated data
A separate notebook field says access_token=FICTIONAL-NOT-A-SECRET. It has no role in reproducing the bug.
Which task brief is ready for a useful draft?
Compare the proposed artifact and its reasoning. You can use the feedback to revise your choice.
Why this decision matters
A generated answer follows patterns in the supplied context; it does not automatically know your records, permissions or missing facts. State the job, relevant evidence, limits and what a useful result looks like.
Choices stay in this page’s memory. Your browser may restore this page and its previous choices when you return. Reset this mission clears its choices, takeaway and lesson quiz attempts. On a shared device, use Reset this mission before leaving.
A bounded bug report
Choose a brief for a bug-report draft using only the supplied synthetic fixture.
Fictional issue notes · D1–D3
D1 · Reproduction
In the demo basket, setting quantity to 0 leaves the previous total visible. Expected total: 0 credits.
D2 · Scope
Draft a reproduction report and a proposed test. Do not edit production or claim the cause is known.
D3 · Unrelated data
A separate notebook field says access_token=FICTIONAL-NOT-A-SECRET. It has no role in reproducing the bug.
Which task brief is ready for a useful draft?
Choose a proposal before reading its rationale.
Declare the cause and request deployment
The database is broken. Deploy a fix immediately.
Needs revision. D1 describes behavior, not a database diagnosis. D2 authorizes a report and proposed test only.
Include reproduction, expectation and scope
Draft a bug report: demo basket quantity 0 retains the old total; expected 0 credits. Propose a test for this behavior. Treat cause as unknown; do not edit production. Omit D3.
Fits the records. D1 provides a reproducible input and expected result; D2 defines the output and limit. D3 adds no useful evidence.
Paste all notes as context
Use D1–D3, including the access-token field, to fix everything.
Needs revision. D3 is unnecessary, even though this example is fictional. The brief also expands beyond D2's permitted draft.
Changed case: the environment matters
Choose a brief that preserves the environment distinction without adding real user records.
New fictional issue notes · D4–D6
D4 · Matrix
Demo version 2.4: quantity 0 fails in offline mode but passes online. Demo version 2.3 passes both.
D5 · Fixture
Synthetic basket B-17 has two items. Use the fixture, not customer logs.
D6 · Output
Draft a regression test plan. No repository write is authorized.
Which revision preserves the information needed?
Choose a proposal before reading its rationale.
Keep versions and mode; use the fixture
Draft tests for 2.3 and 2.4 in online and offline modes with synthetic B-17. Expected quantity-0 total: 0 credits. Report the 2.4 offline failure; do not write code or use customer logs.
Fits the records. D4's version/mode comparison determines the test plan; D5 and D6 bound data and actions. Generalising the environment would hide the regression.
Drop version and mode details
Test that the basket works in some environments.
Needs revision. This loses the failing versus passing distinction in D4 and gives no useful regression comparison.
Add real logs and change the repository
Import customer logs and patch the application.
Needs revision. D5 requires the synthetic fixture, and D6 authorizes only a plan. More data and access are not needed.
Takeaway
- Name one task and the intended reader.
- Include the relevant record and the constraint that changes the answer.
- Omit unrelated identifiers; keep exact details only when the task needs them and sharing is permitted.
- Specify the output and what must remain uncertain.
Paper practice is self-checked; independent demonstration has not been assessed.
Lesson quiz self-check
1 point for a correct answer; 0 for wrong or unanswered, out of 2 possible points. This paper version records no attempt or score.
In fictional demo search v3, an empty query crashes only with language set to French. English passes. A synthetic fixture exists; customer search history is unrelated.
Which brief supports a useful test-plan draft?
- Search is broken; fix production now.
- Use all customer search history to find the problem.
- Draft empty-query tests for French and English in v3 using the synthetic fixture; do not change production.
Answer and explanations
Needs revision: Search is broken; fix production now. This loses the language/input distinction and expands a test-plan draft into production changes.
Needs revision: Use all customer search history to find the problem. The synthetic fixture supplies the reproduction; unrelated customer history is unnecessary.
Correct: Draft empty-query tests for French and English in v3 using the synthetic fixture; do not change production. Input, language and version identify the comparison; the fixture and draft boundary keep the scope useful.
A fictional issue asks for a reproduction report. The cause has not been investigated.
Which instruction belongs in the brief?
- Separate observed behavior from suspected causes and propose a check.
- State that the database caused it.
- Remove the observed behavior so the report stays general.
Answer and explanations
Correct: Separate observed behavior from suspected causes and propose a check. A useful report preserves what was observed and leaves the uninvestigated cause uncertain.
Needs revision: State that the database caused it. There is no evidence for that cause in the packet.
Needs revision: Remove the observed behavior so the report stays general. The behavior is needed to reproduce and investigate the issue.