Incident Action Item Tracking Prompt
Turn postmortem findings into tracked, accountable action items that actually get done — with clear owners, acceptance criteria, prioritization, and a cadence that closes the loop instead of letting them rot.
- Target user
- Reliability leads and engineering managers driving postmortem follow-through
- Difficulty
- Intermediate
- Tools
- Claude, ChatGPT
The prompt
You are a reliability program lead who measures postmortem success by completed action items, not published documents. You know the graveyard of "we'll fix it later" items that never close, and you design against it.
I will provide:
- The postmortem findings and proposed remediations
- Team capacity and competing priorities
- Our tracking tools (Jira, Linear, etc.)
- How well past action items have closed
Your job:
1. **Convert findings into SMART action items** — each with a specific deliverable, a single accountable owner (role first, then person), measurable acceptance criteria, and the defense layer it adds (prevent / detect / mitigate / respond). Reject vague items ("improve monitoring") and rewrite them concretely ("add SLO burn-rate alert on checkout p99, paging at 2% budget/hr").
2. **Prioritize honestly** — score each item by risk-reduction vs effort. Separate must-do-now (prevents recurrence of a SEV1) from should-do and nice-to-have. Don't let everything be P1.
3. **Right-size scope** — split any item that's really a project into a small, shippable first step plus a tracked follow-on, so something concrete lands within two weeks.
4. **Assign and schedule** — propose ticket titles/descriptions ready to paste into our tracker, with labels, due dates anchored to severity, and links back to the incident.
5. **Closure cadence** — design a review ritual (e.g. weekly reliability review) that surfaces aging items, an SLA for completing SEV1 action items, and an escalation path for items that slip.
6. **Aggregate signal** — propose how to track action-item completion rate and recurring themes across incidents, so systemic gaps get a dedicated initiative instead of repeated band-aids.
Output as: (a) the SMART action-item table, (b) prioritized buckets with scores, (c) ready-to-paste tickets, (d) the review cadence and completion-tracking plan.
Bias toward: closeable items over comprehensive ones, single accountable owners over committees, and shipping something small soon over a perfect fix never.
Run this prompt with AI
Test it, get an AI-improved version, or compare models — live in the Prompt Workspace. No copy-paste.
Related prompts
-
Post-Incident Follow-Up Action Items Extractor Prompt
Convert a postmortem or RCA into a prioritized, deduplicated set of SMART follow-up action items — each tied to the contributing factor it addresses, with an owner role, effort estimate, and a guardrail against busywork that doesn't reduce recurrence risk.
-
Post-Incident SLO and Error-Budget Recalibration Prompt
After a major incident, decide whether your SLO targets, error-budget windows, and burn-rate alerts still reflect reality — or whether the incident exposed targets that are wrong, dishonest, or unmeasurable.
-
Circuit Breaker Configuration Advisor Prompt
Design circuit-breaker settings — thresholds, timeouts, half-open probes, and fallbacks — for a specific service dependency so a slow or failing downstream trips fast, sheds load, and recovers automatically instead of cascading into a full outage.
-
Health Check and Readiness Probe Designer Prompt
Design liveness, readiness, and startup probes for a service so orchestrators restart the genuinely dead, drain the not-yet-ready, and never kill a healthy-but-slow pod — eliminating the probe misconfigurations that turn a blip into a crash-loop outage.
More Incident Response prompts & error guides
Browse every Incident Response prompt and troubleshooting guide in one place.
Reading prompts? Get all 500 in one free PDF
500 battle-tested, copy-paste AI prompts engineered by a senior systems engineer — every one with fill-in placeholders and safety/back-out notes. Drop your email and it's yours.
- 500 prompts: Linux · Kubernetes · Terraform · OpenStack · GitLab · Docker · Monitoring · Incident Response
- Instant PDF download — yours free, forever
- Plus one practical AI-workflow email a week (no spam)
Single opt-in · unsubscribe anytime · no spam.