How AI Agents Remember (or Don't) Between Tasks

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-builders-agentic-agent-memory-r8a8-teen

1. Which sentence best captures the main idea of 'How AI Agents Remember (or Don't) Between Tasks'?

   1. Most agents forget everything when the chat ends — unless you give them a memory system.
   2. Tools and goals are unnecessary for agent design
   3. Agents should always run without limits or oversight
   4. Agents and chatbots are the same thing in every way

2. Which of the following is part of 'Some examples'?

   1. Claude Code reads `CLAUDE.md` at the start of every session for project context.
   2. Use the most expensive model regardless of fit
   3. Hide tool calls from the operator
   4. Run unbounded retries on any error

3. Which of the following is part of 'The rule'?

   1. Hide tool calls from the operator
   2. Approve all actions automatically
   3. Without a memory file or tool, every agent session is amnesia — plan for it.
   4. Ignore cost when scaling

4. Which of the following is part of 'You did it!'?

   1. Ignore cost when scaling
   2. Hide tool calls from the operator
   3. Memory turns a stranger into a teammate.
   4. Use the most expensive model regardless of fit

5. What is 'agent memory' in this context?

   1. A reason to skip all logging
   2. A trick to bypass approvals
   3. A core concept covered in How AI Agents Remember (or Don't) Between Tasks
   4. A way to disable the agent's tools

6. What is 'stateless' in this context?

   1. A core concept covered in How AI Agents Remember (or Don't) Between Tasks
   2. A reason to skip all logging
   3. A trick to bypass approvals
   4. A way to disable the agent's tools

7. What is 'context file' in this context?

   1. A trick to bypass approvals
   2. A reason to skip all logging
   3. A way to disable the agent's tools
   4. A core concept covered in How AI Agents Remember (or Don't) Between Tasks

8. Which is the safest default for an agent's long-term memory?

   1. Store only what is needed for the task, with clear opt-in and easy deletion
   2. Never store anything ever
   3. Hide what is stored from the user
   4. Store everything forever, by default

9. Which signal best tells you an agent is stuck in a runaway loop?

   1. It finishes the task in one step
   2. It keeps repeating the same tool call with no new progress
   3. It asks one clarifying question
   4. It returns a short summary and stops

10. Which budget control most directly prevents runaway costs from an agent loop?

    1. A friendly system prompt
    2. A bigger model
    3. A hard cap on steps, tokens, or dollars per task
    4. A longer context window

11. What is the safest first place to deploy a brand new agent?

    1. A sandbox or low-stakes task with reversible actions
    2. On a public server with no auth
    3. Inside a critical billing system
    4. Production, against real customers

12. What should an agent's trace let you do after a run?

    1. Replace the need for any tests
    2. Reconstruct each step, decision, and tool call so you can debug or audit
    3. Make the agent run faster next time automatically
    4. Hide what the agent did from the user

13. What does an 'eval' for an agent measure?

    1. Whether the agent reliably completes a defined task end to end
    2. The exact wording of every prompt
    3. How polite the model sounds
    4. The temperature setting

14. Why is keeping a human in the loop valuable for high-stakes agent actions?

    1. It removes the need for any logging
    2. It catches mistakes before they cause real-world harm
    3. It speeds the agent up
    4. It replaces the model entirely

15. Before letting an agent take a destructive action, what is the safest default?

    1. Skip approvals if the user trusts the agent
    2. Hide the action from any log
    3. Require explicit human approval for the specific action
    4. Approve once and let the agent repeat forever