Blameless Incident Learning Review

The problem

Post-mortems often suffer from hindsight bias, invented causality, and missing evidence that gets "smoothed over" during drafting. Manual incident reviews struggle to maintain a strict audit trail between raw logs and final reports.

What it does

• Builds an evidence ledger with unique source IDs and locators for every material claim.
• Constructs a verified timeline that distinguishes between observations, hypotheses, and causal claims.
• Identifies and preserves conflicting data points in a dedicated conflict log instead of forced reconciliation.
• Generates an action register focused on systemic changes rather than individual blame.
• Flags specific information gaps that require further evidence or human decision-making.

Why this beats prompting it yourself

Generic prompts tend to hallucinate missing timestamps or bridge gaps with plausible-sounding but unverified narratives. This skill enforces a rigorous quality gate system that prioritizes "no false completion," ensuring every conclusion is traceable to your specific logs and telemetry.

Use cases

• Writing blameless post-mortems after production outages or security incidents.
• Synthesizing multiple chat logs, PagerDuty alerts, and CloudWatch metrics into a single source of truth.
• Creating auditable incident reports for compliance or executive review.
• Mapping contributing factors in complex system failures where a single "root cause" is insufficient.

Known limitations

Does not make legal conclusions or attribute individual fault. Requires structured source materials to generate a high-confidence timeline.