Prioritize Requirements

Common Contract (Load First)

When this skill is activated, first load skill-common and apply the shared conventions it defines: canonical section order, frontmatter rules, anti-pattern guidance, and the Review In Mind loop in review-in-mind. This skill adds only domain-specific rules below.

Who

Product Managers, Business Analysts, and AI Agents. The agent uses this skill to objectively rank work items.

What

This skill implements an Analytical Prioritization Matrix. It evaluates a list of requirements, scoring each on relative Business Value, User Value, Cost (effort), and Risk. It calculates a final Priority Score and outputs a sorted backlog.

When

Invoke this skill during sprint planning, backlog grooming, or when deciding MVP scope. Trigger phrases include: "prioritize this list," "what should we build first," "rank these user stories," or "run a prioritization matrix."

Where

Applies to feature lists, user story backlogs, and bug trackers.

Why

Without a strict framework, stakeholders tend to label every requirement as "Critical." By separating the value of a feature from the cost/risk to build it, we can identify "quick wins" (high value, low cost) and avoid "money pits" (low value, high cost).

Inputs

• A list of requirements, features, or user stories.
• (Optional) Stakeholder input on relative weights (e.g., "Cost is more important than Risk right now").

Output (Logical Evidence)

• A markdown table displaying the scores for each requirement.
• A sorted list of requirements from highest priority to lowest.

Optimization Readiness

• Failure Signals: Scores are inconsistent, the matrix falls back to vague High/Medium/Low labels, unknown effort is guessed instead of flagged, or the ranking does not separate quick wins from high-risk investments.
• Evidence To Collect: Scoring tables, weighting choices, stakeholder adjustments, and examples where the ranking was challenged or revised after review.
• Safe Mutation Boundaries: Refine scoring instructions, weighting guidance, output formatting, and review prompts without changing the core numeric prioritization workflow.
• Acceptance Criteria: Accept revisions only if the skill yields a clear numeric ranking, documents weights when used, and calls out unknown estimates instead of masking uncertainty.
• Rejected Revision Handling: Record misleading scoring heuristics, non-numeric fallback patterns, and weak explanations of tradeoffs so they are not reused blindly.
• Transfer Check: Verify the workflow still works for features, requirements, and user stories with and without stakeholder weighting.
• Stop Rule: If there are too few items or too little estimation data to produce a meaningful matrix, stop and tell the user to simplify or gather estimates first.

Constraints (Logical Boundaries)

• Do not use High/Medium/Low. You must use numerical scores (e.g., 1 to 9).
• Do not guess wildly; if a cost is completely unknown, flag it for estimation rather than giving it a random score.

One More Thing

If the provided list has fewer than 3 items, inform the user that a matrix is unnecessary and they should just use simple pairwise comparison.

How (Structural Workflow)

Phase 1: Establish the Scale and Weights

1. Define the scale (usually 1 to 9, where 9 is highest value, highest cost, or highest risk).
2. Assign weights to the categories if the user provided them (e.g., Business Value = 2x, Cost = 1x). Default to 1x for all.

Phase 2: Scoring

3. For each requirement, estimate:
   • Business Value: How much does this help the company (revenue, compliance)?
   • User Value: How much does this improve the user experience?
   • Cost/Effort: How hard is this to build?
   • Risk/Complexity: How likely is this to break or face technical hurdles?

Phase 3: Calculation and Ranking

4. Calculate the Total Value = (Business Value) + (User Value).
5. Calculate the Total Penalty = (Cost) + (Risk).
6. Calculate the Priority Score = (Total Value) / (Total Penalty).
7. Sort the requirements descending by Priority Score.

Resources

• Analytical Matrix Guide

Review In Mind (ReviewInMindGenie)

Execute the common review loop in review-in-mind before delivering.

Review lens for this skill:

• Is every requirement scored on business value, user value, cost, and risk with a calculated priority?
• Are scores traceable to stated criteria rather than hidden judgment?
• Would the sorted backlog be defensible to stakeholders?

Validation

1. Verify that all items in the list were scored.
2. Check the math: ensure the final sorted order matches the calculated Priority Scores.