After-Action Repertoire Review¶

Retrospective review — instantiates Response Repertoire Expansion

A blame-free retrospective that asks, after each handled or mishandled event, whether the system had the right response available — turning recurring gaps into candidate new options and judging whether past additions actually worked.

An After-Action Repertoire Review is a recurring retrospective that inspects the repertoire itself: for each event, did a usable response exist, was it selected, and did it work? It is the discovery-and-feedback engine of the archetype — it both surfaces the recurring unhandled case classes that justify adding a new option and judges whether options already added are actually earning their keep. It builds no responses; it decides which ones are missing or failing. Its whole value is directing repertoire growth from evidence rather than fashion.

Example¶

After a checkout outage at a payments company, the on-call team runs a blameless postmortem. The timeline shows responders had no runbook for the specific failure — a regional cache stampede — and improvised for forty minutes. Read across the last quarter's postmortems, the review notices the same class recurred three times: cache-layer failures with no maintained response. That pattern, not the single incident, is logged as an unhandled case class, with an assessment of why the current repertoire failed — no runbook, no owner, no one trained. The same session revisits a response added last quarter and, from incident data, judges it effective. The output is a prioritized, evidence-backed list of repertoire gaps plus a verdict on prior additions, handed to whoever will build the responses.

How it works¶

The review is retrospective and pattern-seeking. It aggregates across events to separate true recurring gaps from one-offs and routing errors; it diagnoses the cause of failure (a missing skill, tool, authority, or procedure); and it closes the loop on previously-added options using real outcome evidence. It is deliberately blame-free, because the gap reports it depends on only surface when people are not defending themselves. What makes it this mechanism is that it is the archetype's demand signal and its audit at once.

Tuning parameters¶

Cadence and trigger — every incident, periodic, or severity-triggered. More frequent reviews catch more but cost sustained attention.
Recurrence threshold — how many repeats before a gap counts as a maintained-response candidate rather than an anomaly. Too low and the repertoire chases noise.
Blame stance — blameless versus accountability-focused. Blameless surfaces more real gaps but can under-attribute the ownership a fix needs.
Feedback horizon — how long an added option is watched before its effectiveness verdict is called.

When it helps, and when it misleads¶

Its strength is that it keeps repertoire growth demand-driven — grounded in real recurring gaps rather than the latest fashionable tool — and it is the only mechanism that closes the loop on whether past additions worked. Its failure modes are cognitive: hindsight bias makes every handled case look as though it needed a new response, recency over-weights the last dramatic incident, and reviews decay into ritual paperwork nobody reads. The classic misuse is running it to assign blame, which suppresses exactly the gap reports it needs.^[1] The discipline is to keep it blameless and to require a recurrence pattern, not a single incident, before minting a new option.

How it implements the components¶

unhandled_case_class_map — aggregates events into the recurring case classes current responses cannot handle.
response_gap_assessment — diagnoses why the repertoire failed each class: a missing skill, tool, authority, or procedure.
effectiveness_feedback_signal — uses real outcome and incident evidence to judge whether already-added options actually work.

It does not turn a mapped gap into an executable response — specification belongs to Exception Handling Playbook and the other authoring mechanisms — and it does not record the standing baseline of existing capability; that is Competency Matrix Update.

Instantiates: Response Repertoire Expansion — it produces the evidence-backed gap list and effectiveness verdicts that drive every other mechanism.
Sibling mechanisms: Exception Handling Playbook · Controlled Pilot · Cross-Training Program · Competency Matrix Update · Decision Tree Update · Job Aid Checklist · New Service Tier · Runbook Library Update · Scenario Drill · Tool Capability Addition · Triage Protocol Update

References¶

[1] A blameless postmortem reviews an incident to find systemic gaps without attributing fault to individuals; the practice exists precisely because fear of blame is what stops people from reporting the very response gaps this review needs to see. ↩