All products

Write an after-action report with AI

Compare what the team expected with what happened, reconstruct the timeline and give each agreed improvement an owner and a follow-up date.

Write an after-action report with AI while people still remember what they knew during the event. A launch team might discover an hour later that a checkout error had been visible in a different team's logs all along. The after-action-report skill asks why the original decisions made sense with the information available at the time.

Collect an expected outcome, a dated sequence and the effects people actually observed. Keep disagreements visible until the evidence resolves them. The report is most useful when a reader can see what should be repeated, what needs to change and who will check that the change worked.

What it does

  • Compare the original plan with the observed result.
  • Build a timeline from permitted messages, tickets or logs.
  • Distinguish a contributor's observation from a later interpretation.
  • Identify a process or information gap behind a delayed response.
  • Assign a concrete action, owner and review date.

Install and use

  1. Give the agent authorised event notes, participant accounts and the expected impact. Use this workflow after the incident or project, when the team can review it.
  2. Ask, “Why did we notice the registration problem late? Show what was visible at each stage and flag any gaps in the record.”
  3. Review the timeline with participants and confirm the actions with the people who will own them.

FAQ

Source

Made by RampStack Co. Original skill: after-action-report, published on GitHub under the MIT licence.

View on GitHub