Module 02 Knowledge Check
5 questions. Pass mark 4 out of 5. Answer every question before checking the answer key below, then retry after reading the feedback.
1. A goal is finishable when…
- A. it is written clearly
- B. you can name the artefact or decision that means it is done
- C. it mentions the tools to use
- D. it is short
2. The test for keeping an instruction is…
- A. is it polite?
- B. can you name an output it makes impossible?
- C. is it under twenty words?
- D. does it mention the goal?
3. Why do actions need a ceiling separate from steps and tokens?
- A. actions are slower
- B. a run can be cheap in steps and tokens and still perform several irreversible actions
- C. actions cost more tokens
- D. they do not - one ceiling is enough
4. The outcome most designs leave out is…
- A. success
- B. failure
- C. escalate
- D. retry
5. An escalation is actionable when it…
- A. is polite
- B. names what a person would need to resolve it
- C. includes the full trace
- D. happens quickly
---
Answer Key and Explanations
Check these only after attempting every question.
1. B - you can name the artefact or decision that means it is done
'Be thorough' reads like a quality bar and functions as an instruction to keep going.
2. B - can you name an output it makes impossible?
'Try to be accurate' rules out nothing. 'Never estimate a figure that is not written down' rules out a whole class, and you can check it afterwards.
3. B - a run can be cheap in steps and tokens and still perform several irreversible actions
The expensive failure is doing something irreversible too many times, and a step ceiling does not bound that.
4. C - escalate
With only success and failure, a missing input is reported as the agent failing, when a person could resolve it in a minute.
5. B - names what a person would need to resolve it
'Needs a person' is a handover. 'Needs a person: contract not found in the document store' tells them what to do first.
Practical Check
Apply this module to your own work: complete the module activity for *Tasks, Goals, Instructions and Stopping Rules*, then write one sentence naming what your result shows and one naming what it does not.
Strong Answer Pattern
A strong answer names the task, the evidence used, the check performed, and the remaining limitation. It avoids "proved", "guaranteed", or "always" unless the evidence genuinely supports it.