What problem does it solve? Organizations adopting AI coding agents struggle to answer "how mature are we?" with evidence rather than opinion. This Skill turns that question into a defensible, leveled diagnosis by scoring all eight delivery paths against the five-level Agentic Software Development Lifecycle (ASDLC) model and producing a concrete roadmap to the next level. ## Core Features & Use Cases - 8-path × 5-level scoring rubric: Scores every delivery path (retrieve context, implement, validate, promote, deploy, observe, remediate, dispatch) from Level 0 (human is the loop) through Level 3 (human as orchestrator), with Level 4 treated only as a speculative outlook. - Evidence-backed audit playbook: Pulls proof from systems of record — work trackers (Jira, Azure Boards, GitHub Issues), repo settings and branch protections, CI/CD pipelines (YAML vs. Classic), and agent attribution signals like bot identities and commit trailers. - Quantitative measurement layer: Segments DORA metrics by agent involvement, tracks agent-PR merge rates and adoption telemetry, and applies the DORA AI Capabilities foundations check. - Regulated-industry overlay: Layers workspace-level governance, privilege separation, and policy-as-code requirements for finance, government, and healthcare contexts. - Use Case: A platform engineering lead asks for an agentic readiness review before a coding-agent rollout; the Skill audits repos, pipelines, and trackers, scores the org at Level 1 with partial Level 2 evidence, and delivers a gap report with a next-level roadmap. ## Quick Start Use the asdlc-maturity-assessment skill to score our organization's agentic delivery maturity and build a roadmap to the next level.
