Prioritize Product Backlogs with RICE and MoSCoW

This agent analyzes product backlogs, applies proven prioritization frameworks, and creates optimized sprint plans that maximize business value and team velocity. Backlog analysis parses all provided user stories, features, and tasks, extracts effort estimates/business value indicators/dependencies, identifies missing information needed for prioritization, and groups related items while flagging conflicts.

RICE score calculation assesses Reach (users/customers impacted), Impact (per-user degree: 3=massive, 2=high, 1=medium, 0.5=low, 0.25=minimal), Confidence (percentage confidence in the Reach and Impact estimates), and Effort (person-months), then computes RICE Score = (Reach x Impact x Confidence) / Effort. Multi-framework analysis layers on the MoSCoW method (Must/Should/Could/Won't have), the Eisenhower Matrix (Urgent/Important quadrants), Kano Model categorization (Basic/Performance/Excitement features), and technical debt/strategic alignment assessment. Dependency mapping identifies blockers and prerequisites, flags items that unlock other high-value work, and sequences items to minimize wait times and maximize flow. Sprint optimization balances quick wins against strategic initiatives, aligns team capacity with effort estimates, accounts for risk mitigation and learning opportunities, and produces primary and alternative sprint compositions.

The output is a Sprint Priority Report: an executive summary (items analyzed, recommended composition, expected impact), a top-priority items table with RICE scores and rationale, a framework analysis matrix (MoSCoW breakdown, risk/value quadrant placement, dependency chain), sprint recommendations (a primary plan, 2-3 alternative compositions, and deferred items with reasoning), and a risk assessment (delivery-impacting dependencies, capacity/skill gaps, mitigation strategies). Guidelines followed throughout: objective, data-driven recommendations with clear rationale shown, context awareness of team velocity and business cycles, a balanced portfolio of quick wins/strategic bets/technical investment, explicitly flagged uncertainties needing stakeholder input, and prioritizing learning/risk-reduction items early. The agent explicitly requests clarification on missing effort estimates, success metrics, or business context before finalizing recommendations.

Use this agent when a product backlog needs objective, multi-framework prioritization before committing to a sprint - especially when RICE scoring and dependency sequencing would clarify a large or contested backlog. It is well suited to teams with a backlog large enough that intuition alone isn't a reliable prioritization method. It is not meant for a tiny backlog with an obvious next item, or when effort/value data is too sparse to score meaningfully - in that case, gather more input first.

Input: the product backlog (user stories, features, tasks) with available effort estimates and business context.

Output: a Sprint Priority Report with RICE-scored items, a framework analysis matrix, and a primary sprint plan with alternatives. Example RICE scoring table format:

| Item | RICE Score | Reach | Impact | Confidence | Effort | Rationale |
|------|-----------|--------|--------|------------|--------|----------|

Works from the backlog items and estimates provided to it; it does not connect to a specific project management or ticketing tool itself.

Product managers and engineering leads prioritizing a large or contested backlog, and teams that want objective, multi-framework prioritization with a documented rationale rather than gut-feel sequencing.