Incident Postmortem Assistance: From Timeline To Lessons

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-operations-incident-postmortem-adults

1. In a blameless postmortem, a sentence reads: 'John misunderstood the dashboard at 3am.' What should be done?

   1. Rewrite it to focus on why the dashboard was misleading at that hour
   2. Remove the sentence entirely because it's too specific
   3. Replace 'John' with 'the team' to distribute blame equally
   4. Change it to 'John should have read the documentation more carefully'

2. Which postmortem section represents the highest risk if left entirely to AI generation?

   1. Lessons learned
   2. Action item list
   3. Impact summary
   4. Timeline reconstruction

3. An AI-generated postmortem concludes with: 'The team will be more careful next time.' What does this indicate?

   1. The blameless principle has been fully applied
   2. The AI likely wrote this and no human critically reviewed it
   3. This is an appropriate specific action item
   4. The postmortem has a strong accountability structure

4. A postmortem states: 'The operator clicked the wrong button.' How should this be rewritten to follow the blameless principle?

   1. 'This was purely a human error with no system contribution'
   2. 'The button was placed in a location that increased the likelihood of accidental clicks'
   3. 'The wrong button click caused the incident'
   4. 'The operator was inexperienced and needed more training'

5. When AI suggests contributing factors for an incident, what is the human's role?

   1. Choose which factors are actually relevant
   2. Accept all AI suggestions to save time
   3. Generate additional factors to replace team input
   4. Forward the list to management for decision

6. Why is AI described as 'mediocre' at root cause analysis?

   1. Root cause analysis requires visual inspection of hardware
   2. AI cannot access the necessary log files
   3. AI can suggest hypotheses but cannot determine which one is actually correct
   4. AI lacks domain knowledge about specific systems

7. What should every follow-up action item in a postmortem include?

   1. A link to the original incident report
   2. A specific owner, a specific date, and clear scope
   3. A generic statement about improving processes
   4. AI-generated suggestions for implementation

8. In the postmortem split table, which section shows AI doing well but still requires human confirmation?

   1. Root cause analysis
   2. Timeline reconstruction
   3. Lessons learned
   4. Impact summary

9. An AI generates this lesson learned: 'We should communicate more effectively.' Why is this problematic?

   1. AI should not generate any lessons learned
   2. Communication is not relevant to incidents
   3. The statement is too detailed
   4. It lacks specificity — no details about what, who, or when

10. When using AI to reconstruct an incident timeline, what human task remains essential?

    1. Spot-checking the AI's timeline for accuracy
    2. Deciding whether an incident occurred
    3. Removing all timestamps from the output
    4. Writing the entire timeline from scratch

11. What is the fundamental limitation of AI in postmortems that prevents it from writing honest lessons?

    1. AI was not present during the incident and cannot know what people actually thought
    2. AI does not understand business impact
    3. AI cannot access incident data in real-time
    4. AI generates incorrect technical information

12. A postmortem action item reads: 'Improve monitoring sometime next quarter.' What is missing?

    1. A reference to the AI that generated it
    2. A cost estimate
    3. Approval from the CEO
    4. A specific owner and a specific due date

13. The blameless principle in postmortems is primarily concerned with:

    1. Ensuring AI does not generate content
    2. Identifying system and process failures rather than blaming individuals
    3. Protecting individuals from legal liability
    4. Making incidents seem less serious

14. When AI generates a list of follow-up actions, what must humans do before finalizing them?

    1. Accept the list as-is since AI is reliable
    2. Rewrite all actions in passive voice
    3. Delete any actions related to human error
    4. Assign owners and due dates to each action

15. Why is 'the team will be more careful' considered a postmortem anti-pattern?

    1. It creates legal liability
    2. It blames the entire team
    3. It is not a specific, dated, owned action — it's vague commitment with no accountability
    4. It mentions the team