What problem does it solve? Turning a plan, spec, or conversation into actionable work items is error-prone: slices are often horizontal (one layer at a time), dependencies between tasks go undeclared, and tickets lack verifiable acceptance criteria. This Skill decomposes work into vertical tracer-bullet tickets, each with explicit blocking edges, a risk tier, and machine-checkable acceptance criteria, then publishes them to a tracker. ## Core Features & Use Cases - Vertical slice decomposition: Splits work into narrow end-to-end slices that cut through schema, API, UI, and tests, each demoable on its own and sized for a single context window. - Dependency graph declaration: Every ticket declares which tickets block it, published as text edges in local Markdown files or as native blocking links on GitHub, Linear, or similar trackers. - Risk tier and quality gate tagging: Each ticket carries a risk tier computed from the project's high_risk_paths configuration and a gate mode its verify step must satisfy. - Wide refactor handling: Sequences large mechanical changes as expand-migrate-contract ticket chains instead of forcing them into a single vertical slice. - Use Case: After drafting a feature spec, ask the agent to break it into tickets; it proposes a numbered breakdown with blocking edges, quizzes you on granularity, then publishes one file or issue per ticket in dependency order. ## Quick Start Ask the agent to break the current plan or a referenced spec into tracer-bullet tickets and publish them to the configured tracker.
