prioritize

Prioritization

Score and rank features or initiatives so the most valuable, strategic work rises to the top. The default framework is RICE — opinionated enough to force a decision, flexible enough to reflect your strategy.

Step 0: Context Scan (Silent)

Before asking anything, scan for strategic context:

cat STRATEGY.md 2>/dev/null
find . -maxdepth 2 -name "OKRs-*.md" 2>/dev/null | sort | tail -1 | xargs cat 2>/dev/null
find . -maxdepth 3 -name "FR-*.md" -o -name "feature-request-*.md" 2>/dev/null | head -20 | xargs cat 2>/dev/null
find . -maxdepth 2 -name "prioritization-*.md" 2>/dev/null | sort | tail -1 | xargs cat 2>/dev/null

Use what you find to:

• STRATEGY.md found: pull the strategic pillars and use them to assess strategic fit for each item. Note which pillar each item maps to.
• OKRs found: use current Key Results to calibrate Impact scores — items that move an active KR score higher.
• FR files found: when an item in the input matches a known FR, read the FR for reach estimates, user context, and scope. Do not ask the user to re-describe items that are already specced.
• Prior prioritization found: surface it and ask if the user wants to update it or start fresh.
• Nothing found: proceed with user-supplied context only.

Do not mention the scan.

Step 1: Input

Accept items in any format — a numbered list, bullet points, ticket IDs, FR filenames, or a paragraph. Extract the item name and any context provided.

If the input is sparse (item name only, no context), ask for the missing inputs required to score that item rather than scoring blind.

If the user specifies a framework (Agentic RICE, ICE, MoSCoW, custom), switch to it. See Alternative Frameworks below.

If the user mentions agentic workflows, AI agents, or autonomous execution, suggest Agentic RICE and explain the difference before proceeding.

Step 2: Score Each Item

Default framework is RICE. Score every item, then rank by RICE score descending.