What problem does it solve? Teams need coarse, order-of-magnitude effort estimates for stories and epics before sprint planning, but producing consistent ROM baselines with explicit assumptions is often skipped or done ad hoc. ## Core Features & Use Cases - T-shirt Sizing Baseline: Maps stories to XXS through XXL sizes based on human-equivalent effort days, with XXL acting as a split gate that routes oversized items to decomposition. - Ranged Estimates: Produces low, expected, and high effort ranges with explicit assumptions and exclusions instead of false-precision single numbers. - Local Backlog Recording: Records the ROM baseline against the repo-resident backlog task in .setup-agents/state/tasks.jsonl, with Atlassian MCP as an optional mirror only. - Use Case: A BA or PM refines an epic, estimates each story independently, rolls the results up to the epic, and records the ROM table so planning and prioritization can consume it. ## Quick Start Ask the assistant to generate a ROM estimate for a refined story or epic and record the baseline against the local backlog task.
