What problem does it solve?
Large or vague TAPD requirements are hard to estimate, schedule, and deliver. This Skill analyzes a requirement document, splits oversized stories into right-sized sub-requirements, and produces consistent effort, size, and business-value scores before creating child stories in TAPD.
Core Features & Use Cases
- Requirement Breakdown: Splits stories using MECE principles, MVP-first slicing, FURPS+ NFR scanning, and interface-contract-first planning with parallel wave编排, then validates coverage with a traceability matrix and 6 consistency checks.
- Three-Dimensional Scoring: Estimates effort via PERT three-point estimation with type-specific baselines (development, ops, documentation), computes a Fibonacci size from 5 scale dimensions, and calculates a RICE business-value score written back to TAPD fields.
- Guardrails & Ticket Creation: Enforces a 24-person-hour cap and size thresholds with up to 3 re-split callbacks, then creates child stories through a four-stage protocol (pre-check, create, post-verify, auto-fix).
- Use Case: Given a TAPD story ID, the Skill fetches the story, splits it into 3 sub-requirements saved under docs/reqs/, scores each one, and creates approved child stories in TAPD with effort, size, and business_value filled in.
Quick Start
Ask the agent to evaluate TAPD requirement 12345 and it will fetch the story, split and score it, then create the sub-requirement tickets.
