Portfolio Prioritizer & WIP Guardrails
Turn "everything is priority 1" into a ranked queue with explicit decision criteria
Project prioritization is hard to embed, and teams suffer context-switching fatigue when too many streams run in parallel. This recipe implements a scoring model with WIP limits and a cadence for re-ranking that prevents chaos.
INGREDIENTS
PROMPT
Create a skill called "Portfolio Prioritizer & WIP Guardrails". Input: project intake list + constraints. Output: ranked portfolio + WIP policy + governance cadence. Required behavior: - Propose a scoring model if none exists; allow me to adjust weights. - Produce a ranked list with rationale per item. - Recommend WIP limits and a cut line. - Generate a portfolio review agenda and a decision packet for executives. Guardrails: - Do not pretend scoring is objective; always show weights and assumptions. - If data is missing for a project, mark it as "needs discovery" instead of guessing.
How It Works
Feed in your project intake list and constraints. The recipe proposes a scoring model
(or uses yours), scores and ranks every project, applies WIP limits to determine a cut
line, and produces a decision pack for executive sign-off — including the cost of
re-prioritizing mid-flight.
What You Get
- Ranked backlog: score breakdown per project with rationale
- WIP limit recommendation: portfolio-level and per-team
- Cut line: what won't start and why
- Change cost note: what it costs to re-prioritize mid-flight
- Executive decision pack: top 10 projects with justification
- Governance cadence: portfolio review agenda (monthly/biweekly)
Setup Steps
- List project requests: title, requester, value hypothesis
- Define or accept suggested scoring criteria (value, risk, effort, strategic fit, etc.)
- Provide resource constraints and WIP capacity
- Review the ranked list and decision pack
Tips
- Scoring is transparent — weights and assumptions are always visible
- Scoring models: simple 2×2, weighted scorecard, or MoSCoW
- If data is missing for a project, it's marked "needs discovery" instead of guessed
- Swap policy: require trade-off, or allow emergency-only swaps
- No ad-hoc priority swaps without a change request — that's the whole point