• Practice
  • About 25 min
  • 15 practice questions

Final self-assessment: AI agents and automation

Test yourself on the whole course: 15 questions with answers explained.

If something looks wrong, please tell us.

Attempt all 15 questions in one sitting of about 25 minutes before looking at any answer. The questions mix ideas from all five lessons and are harder than the lesson checks. When you finish, mark your answers and read every explanation, including those for questions you got right.

Final self-assessment

  1. 1. Each morning an office receives 30 differently worded citizen emails, and each must be routed to the right department. Which statement is most accurate?

    1. A fixed script is ideal because the emails are worded differently
    2. Reading varied text needs judgement, so a model can do the routing step, but its choices should be checked and logged
    3. No software can help with this kind of task
    4. An agent should also reply to each citizen on its own to save time
    Show answer

    Answer: B. Varied wording is where a model helps; a fixed script cannot read free text. But the routing is a judgement that can be wrong, so it is logged and checked, and replying is an outward action that needs approval.

  2. 2. Which sequence correctly shows the agent loop?

    1. Act -> goal -> observe -> plan
    2. Plan -> observe -> goal -> act
    3. Observe -> act -> plan -> goal
    4. Goal -> plan -> act with a tool -> observe -> repeat or stop
    Show answer

    Answer: D. A person sets the goal, the model plans, asks for a tool, observes the result and decides whether to continue or stop.

  3. 3. An agent processing 10 files misreads a date in the first file and uses that date in every later step. This shows that agents are weak at:

    1. Long chains of steps, where an early error is carried forward
    2. Reading any kind of file
    3. Running on a schedule
    4. Using tools of any kind
    Show answer

    Answer: A. In a long chain, each step builds on earlier ones, so one early mistake spreads. That is why results are checked rather than trusted.

  4. 4. An API returns status code 401. What does it most likely mean?

    1. The request succeeded
    2. The requested item does not exist
    3. The caller is not authorised, for example because its key is missing or wrong
    4. The server wants the model to make a tool call
    Show answer

    Answer: C. 401 means not authorised. Success is 200 and not found is 404.

  5. 5. A batch API returns {"batch": "Morning", "students": ["Imran", "Kavya", "Sunil", "Rekha"], "room": 6}. How many students are listed, and what is the room number?

    1. 3 students; room 6
    2. 4 students; room 6
    3. 4 students; room 4
    4. 6 students; room 4
    Show answer

    Answer: B. The square brackets hold a list of four names, so there are 4 students. The value of "room" is the number 6.

  6. 6. A model asks for the tool bookSeat with test 31 and student 1042. What should happen next in a safe design?

    1. The model connects to the booking server by itself
    2. The booking is assumed done and the model reports success
    3. The booking API key is added to the model's prompt
    4. The software checks the call is allowed - and, since it is a write, may ask a person - then runs it with its own stored key and returns the result
    Show answer

    Answer: D. The model only proposes. The software checks, may require approval for a write, runs the request with a key the model never sees, and returns the real result.

  7. 7. A long task fills the agent's context window, and it begins ignoring a rule given at the very start. What is the most likely cause and fix?

    1. Older material was dropped or squeezed out; keep the context lean and restate key rules
    2. The training data was deleted; retrain the model
    3. The API key expired; renew it
    4. Long-term memory is too small; store more notes about users
    Show answer

    Answer: A. The context window is limited, so early instructions can be lost as tool results pile up. Passing only what each step needs and restating rules fixes it.

  8. 8. Which description fits retrieval?

    1. The model automatically remembers every past conversation
    2. Copying the whole database into every prompt
    3. Searching a store of notes or documents and placing only the relevant entries in the context window
    4. Downloading the model's training data
    Show answer

    Answer: C. Retrieval finds the few relevant entries in long-term memory and brings them into the context window for the current task.

  9. 9. A help agent serves many students from one memory store. Which rule matters most for privacy?

    1. Keep every note for ever so nothing is lost
    2. Store OTPs so students need not retype them
    3. Let the model decide on its own what is private
    4. Keep each person's notes separate so they are never retrieved into another person's conversation
    Show answer

    Answer: D. Separating each person's notes stops one student's data leaking into another's answers. OTPs must never be stored, and notes should expire.

  10. 10. Which design choice best limits the damage if an agent is tricked by a prompt injection?

    1. A longer, more detailed system prompt
    2. Least privilege, with the agent holding only the permissions its task needs and write actions needing approval
    3. A faster model
    4. Hiding the logs from staff
    Show answer

    Answer: B. Models cannot be relied on to spot every injection, so the design limits what a tricked agent can do. Fewer permissions and approval for writes cap the damage.

  11. 11. An approval screen for an outgoing message should show:

    1. The exact message and the exact recipients that will be used
    2. Only the agent's one-line description of what it plans to send
    3. The API key that will be used to send it
    4. Nothing - approval should be one automatic click
    Show answer

    Answer: A. The approver must see the real action, because the agent's description of it may be wrong. Keys are never shown.

  12. 12. Which log entry is badly designed?

    1. 10:02 sendMessage - 48 recipients - approved by the duty teacher
    2. 10:01 readDocument - notice file - 2 pages read
    3. 10:03 sendMessage - logged together with the full messaging API key
    4. 10:00 fetchPage - saved official address - page changed
    Show answer

    Answer: C. Logs must never contain secrets. A key written into a log can be read by anyone who sees the log. The other entries record time, tool, inputs, results and approval, as they should.

  13. 13. In the notice workflow, why is new-notice detection done by a plain comparison script rather than by the model?

    1. Because models cannot handle text of any kind
    2. Because scripts are more creative than models
    3. Because a model would need the messaging key
    4. Because spotting a new link in a list is a fixed, well-defined job, so a script does it more predictably and cheaply
    Show answer

    Answer: D. Use a script where the steps are fully known and the model only where reading or judgement is needed - here, summarising and drafting.

  14. 14. A teacher approves a draft after correcting a wrong date. What should the log keep?

    1. Only the final message, so the record looks clean
    2. The original draft, the edit, who approved and when
    3. Nothing, because a person was involved
    4. The teacher's password, to prove who approved
    Show answer

    Answer: B. Keeping the original and the edit shows how often the model errs and who took responsibility. Passwords never go into logs.

  15. 15. Which action should the recruitment-notice agent never be allowed to take?

    1. Fetching the saved official page
    2. Saving a draft reminder for review
    3. Filling in and submitting an application form for a student
    4. Writing a log entry for each step
    Show answer

    Answer: C. Applying is the student's own decision and involves personal data and possibly fees. The agent only reads, drafts and logs; anything outward waits for approval.

Revisit the lessons for any question you missed: questions 1-3, lesson 1 (what agents are); questions 4-6, lesson 2 (tools and APIs); questions 7-9, lesson 3 (workflows and memory); questions 10-12, lesson 4 (human approval and safety); questions 13-15, lesson 5 (the end-to-end workflow).

Sign in to keep track of the lessons you finish.

Privacy · Terms · Refunds · Delivery