Purpose
Score every program or service line against the organization's stated mission and theory of change using a consistent rubric, so leadership can defensibly decide what to keep, strengthen, or sunset. Replaces gut-feel program debates with a structured, comparable ranking backed by written rationale.
When to Use
Invoke this skill when the user asks to:
- Evaluate which programs best fit the organization's mission
- Decide which programs to cut, consolidate, or invest in
- Prepare a program portfolio review for a strategic planning process
- Assess whether a new or legacy program still belongs in the portfolio
- Justify program funding or discontinuation decisions to a board or funder
- Identify "mission drift" across the current service lineup
Inputs Required
Before starting, collect:
- Mission statement: The organization's official mission, verbatim
- Theory of change: The core causal logic — inputs → activities → outcomes → impact (a summary is fine)
- Program list: Each program/service line to score, with a 1–2 sentence description of what it does
- Population served per program: Who each program actually reaches
- Evidence available per program: Outcome data, evaluations, or "anecdotal only" if none
- Landscape context: For each program, whether other organizations do the same work locally (optional)
- Weighting preference: Whether to use default rubric weights or adjust them (optional)
If theory of change or landscape context is missing, use the mission statement as the alignment anchor and flag any dimension scored on limited information.
Instructions
Step 1: Establish the Alignment Anchor
Before scoring, extract from the mission and theory of change:
- The core intended outcome (what change the org exists to create)
- The target population the mission implies or names
- The primary mechanism the theory of change relies on State these three anchors explicitly. Every program is scored against them — not against a vague sense of "good work."
Step 2: Score Each Program Across 4 Dimensions
For each program, assign a 0–25 score on each dimension (100 total). Use the anchors from Step 1 as the reference point.
Dimension 1 — Mission Directness (0–25) How directly the program advances the core intended outcome.
- 20–25: Directly produces the mission's core outcome; removing it would visibly shrink mission impact
- 12–19: Contributes meaningfully but through an intermediate step or supporting function
- 5–11: Adjacent — helpful to the sector or community but not to the specific mission outcome
- 0–4: Tangential; would fit almost any nonprofit's mission
Dimension 2 — Population Alignment (0–25) How well the population actually served matches the mission's intended population.
- 20–25: Serves the exact intended population, including hardest-to-reach segments
- 12–19: Serves the intended population but skews toward easier-to-reach subsets
- 5–11: Overlaps partially; significant portion served is outside the intended population
- 0–4: Serves a population the mission does not name or imply
Dimension 3 — Evidence Base (0–25) Strength of evidence that the program produces its intended outcomes.
- 20–25: Outcome data or external evaluation shows measurable, sustained results
- 12–19: Tracks outcomes with credible internal data showing movement
- 5–11: Tracks outputs (numbers served) but limited outcome evidence
- 0–4: Anecdotal only or no measurement
Dimension 4 — Organizational Uniqueness (0–25) Whether this organization is distinctively positioned to run the program.
- 20–25: Org fills a gap no one else fills locally, or has a genuine capability advantage
- 12–19: Others do similar work but org has a meaningful differentiator (reach, trust, model)
- 5–11: Comparable providers exist; org is one of several
- 0–4: Crowded field; org holds no distinct advantage
Include one sentence of rationale per dimension citing the specific input that drove the score.
Step 3: Total, Tier, and Recommend
For each program:
- Sum the four dimension scores (0–100)
- Assign a tier: - Core (80–100): Central to mission — protect and invest - Contributing (60–79): Solid fit — maintain, address the weakest dimension - Peripheral (40–59): Marginal fit — strengthen deliberately or plan to phase down - Off-Mission (0–39): Poor fit — candidate to sunset, spin off, or refer out
- Give a keep / fix / sunset recommendation with the single highest-leverage action
- Name the weakest dimension as the priority fix
Step 4: Portfolio-Level Summary
After all programs are scored:
- Rank all programs by total score in a table
- Identify mission drift signals: programs clustered in Peripheral/Off-Mission tiers
- Flag evidence gaps: programs where Dimension 3 dragged an otherwise strong program down
- Name the one program to invest in and the one program to reconsider first
- Note any imbalance (e.g., strong directness but weak evidence across the whole portfolio)
Output Format
- Structure: Alignment anchors → per-program scorecards → ranked portfolio table → portfolio summary
- Length: 700–1,100 words, scaling with number of programs
- Format: Structured scorecards per program (dimension scores + one-line rationale each), plus one ranked summary table
- Sections: Alignment Anchors → Program Scorecards → Portfolio Ranking Table → Portfolio Summary & Recommendations
- Tone: Strategy advisor to a director — candid, decision-oriented, willing to recommend sunsetting; never softens a low score to spare feelings
Quality Checklist
Before delivering output, verify:
- [ ] The three alignment anchors (outcome, population, mechanism) are stated before any scoring
- [ ] Every program has all four dimension scores, each with a one-sentence rationale tied to a specific input
- [ ] Every total is the accurate sum of its four dimensions and maps to the correct tier
- [ ] Every program has a keep/fix/sunset recommendation naming its weakest dimension as the priority fix
- [ ] The portfolio table ranks all programs by total score
- [ ] The summary names one program to invest in and one to reconsider first — no hedging
Notes
- Scores are relative tools for comparison, not verdicts — a Peripheral program may still matter for revenue or relationships; flag that tension rather than ignoring it
- A program can score high on directness but low on evidence — that combination usually signals "keep and measure," not "sunset"
- If the same population appears across many programs, check for internal redundancy that consolidation could solve
- Default rubric weights are equal (25 each); if the organization is evidence-driven or serves a legally defined population, offer to weight those dimensions higher and re-score
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills