What problem does it solve? Teams running the ICEA workflow lack an objective way to know whether the process is actually improving code quality and reducing rework. This Skill queries Azure DevOps directly after each sprint and produces a KPI report covering both the benefit side (ICEA compliance rate, PR rejection rate, rework hours) and the cost side (time-in-gate, critic retries, gate overrides, dismissal churn, manifest predictability), so retrospectives run on data instead of impressions. ## Core Features & Use Cases - Three benefit KPIs from ADO REST API: ICEA compliance rate (tickets reaching Active with Intent and Acceptance sections), PR rejection rate (abandoned vs completed PRs), and rework hours (Completed Work on sprint-created Bugs, with a 2h-per-bug fallback estimate). - Five cost-side metrics from local artifacts: time-in-gate from git history, critic retries from session logs, gate overrides from commit messages, dismissal churn from ledgers, and manifest predictability from recorded accuracy lines. - Governance reconciliation: pulls ADO-verified PR approver identities into the audit trail and flags ICEAs marked Approved whose PR has no approving lead vote. - Use Case: After Sprint 5 closes, a delivery lead runs the skill with the sprint name and capacity, receives a threshold-rated report (✅/⚠️/❌) with detail lists of non-compliant tickets, abandoned PRs, and bugs, plus a net-value trend versus the previous sprint. ## Quick Start Ask the assistant to run sprint metrics for Sprint 5 with a capacity of 320 hours to generate the KPI report from Azure DevOps.
