What problem does it solve? Postmortems and post-release metrics often end up as documents nobody rereads, so incidents repeat and failed hypotheses never change decisions. This Skill closes the spec-driven loop by turning that evidence into concrete, owned actions instead of archived reports. ## Core Features & Use Cases - Hypothesis Verdict: Confirms or refutes the hypothesis defined during discovery, treating a failed experiment as valid learning rather than an ambiguous outcome. - Corrective Actions with Owners: Produces specific changes to prevent incident recurrence or capitalize on what worked, each assigned to an owner. - Technical Debt and Backlog Capture: Records shortcuts taken under pressure (with HACK markers and upgrade triggers) and generates prioritized backlog items ready to re-enter discovery or specification. - Use Case: After a production incident is resolved, invoke the Skill with the incident timeline to produce a confirmed root-cause decision, three owned corrective actions, and two new prioritized backlog items. ## Quick Start Run /learn with the post-incident timeline or the measured post-release success metric to generate decisions, corrective actions, and backlog items.
