Purpose
Score and rank a defined set of infrastructure delivery options against a weighted, multi-criteria framework — producing a defensible options comparison matrix, a ranked recommendation, and a sensitivity check that shows how the ranking holds under different weighting assumptions. Solves the problem of subjective, unauditable option selection that fails scrutiny at business case gateways.
When to Use
Invoke this skill when the user asks to:
- Appraise a set of infrastructure delivery or investment options against defined criteria
- Build the options assessment section of a Strategic or Outline Business Case
- Compare do-nothing, do-minimum, reference project, and PPP/alternative delivery routes
- Justify shortlisting or eliminating options for a gateway review or approval board
- Produce a weighted scoring matrix that survives audit or independent assurance
- Test whether a preferred option remains preferred under different criteria weightings
Inputs Required
Before starting, collect:
- Project description: 3–5 sentences on the asset, service outcome, and geographic scope
- Options to evaluate: The named list of delivery options (minimum 3, typically 4–6)
- Strategic objectives: The 2–4 outcomes the project must deliver
- Evaluation criteria and weights: If provided, use them; if not, apply the default framework below
- Cost data available: Capital cost, whole-life cost, or order-of-magnitude estimates per option
- Key constraints: Funding envelope, delivery deadline, statutory or planning constraints
- Risk appetite / assurance context: Approval body, gateway stage, and any political sensitivities
If weights are not provided, apply the default framework (Cost 25%, Risk 20%, Deliverability 20%, ESG 15%, Strategic Fit 20%) and state the assumption explicitly.
Instructions
Step 1: Confirm the Option Set and Baseline
Establish the appraisal boundary before scoring:
- Confirm every option is genuinely viable — eliminate any that fail a threshold constraint (funding, statutory, safety) and record why in a short exclusion note rather than scoring them
- Confirm the do-nothing / do-minimum baseline is present — every appraisal needs a counterfactual to score value against. If absent, add do-nothing as the reference case
- Ensure options are described at comparable definition — do not score a fully-specified reference project against a one-line PPP concept without flagging the asymmetry
Step 2: Establish the Weighted Criteria Framework
Set or confirm the scoring criteria. Use these five default criteria with sub-factors:
Cost (default 25%) — capital cost, whole-life/operating cost, funding certainty, affordability against the envelope
Risk (default 20%) — delivery risk, cost/schedule overrun exposure, risk transfer effectiveness, residual risk retained by the sponsor
Deliverability (default 20%) — market capacity and appetite, procurement complexity, planning/consenting path, timeline to service, dependency on scarce skills
ESG (default 15%) — carbon (embodied and operational), environmental impact and mitigation, social value/community outcomes, resilience and adaptation
Strategic Fit (default 20%) — alignment to strategic objectives, flexibility/optionality for future change, alignment to policy and stakeholder priorities
- Confirm weights sum to 100%
- If the sponsor supplies bespoke criteria, adopt them and map each to a weight
- State any criterion that cannot be assessed due to missing data and score it conservatively with a flag
Step 3: Score Each Option Against Each Criterion
Apply a consistent 1–5 scale per criterion, where 5 is strongest:
- 5 = clearly superior / meets objective with margin
- 4 = strong / meets objective
- 3 = acceptable / meets threshold
- 2 = weak / partially meets
- 1 = poor / fails or materially compromised
For every score, write a one-line justification tied to evidence or a stated assumption. Do not award scores without rationale — an unexplained score is not defensible at gateway.
Compute the weighted score per criterion (raw score × weight) and a total weighted score per option out of 5.00.
Step 4: Build the Options Comparison Matrix
Assemble a table with options as columns and criteria as rows:
- Show raw score and weighted contribution per cell
- Show the total weighted score and rank per option in a summary row
- Include a short qualitative "key strength / key weakness" line beneath each option's total
Step 5: Sensitivity Analysis
Test the robustness of the ranking:
- Re-run the ranking under 2–3 alternative weighting scenarios (e.g., cost-led, risk-led, ESG-led)
- Identify whether the preferred option switches under any plausible weighting — a preferred option that only wins under one weighting is fragile and must be flagged
- Note the criteria to which the outcome is most sensitive
Step 6: Recommendation and Next Steps
- State the ranked preferred option and the runner-up
- Give a 3–4 sentence rationale grounded in the matrix, not restated opinion
- Note the key risks or evidence gaps the sponsor must close before the next gateway
- Flag any option worth carrying forward as a fallback
Output Format
- Structure: Narrative preamble → weighted criteria framework → options comparison matrix (table) → sensitivity analysis → ranked recommendation
- Length: 900–1,300 words plus the matrix table
- Format: Structured document with one central comparison table; scoring justifications as concise bullets
- Sections: Option Set & Baseline → Criteria & Weights → Scoring Justifications → Comparison Matrix → Sensitivity Analysis → Recommendation & Next Steps
- Tone: Infrastructure advisory director — evidence-led, neutral in scoring, decisive in the recommendation; written to withstand independent assurance
Quality Checklist
Before delivering output, verify:
- [ ] A do-nothing or do-minimum baseline is included among the scored options
- [ ] Criteria weights sum to exactly 100% and are stated
- [ ] Every score has a one-line justification — no unexplained numbers
- [ ] The comparison matrix shows both raw scores and weighted contributions, with totals and ranks
- [ ] Sensitivity analysis re-ranks under at least 2 alternative weightings and flags any option switch
- [ ] The recommendation names a single preferred option and does not hedge
Notes
- The most common failure at gateway is scores without evidence — a matrix is only defensible if each cell traces to a stated basis
- If cost data is order-of-magnitude only, score the Cost criterion on a relative basis across options and flag the estimate class rather than inventing precision
- A preferred option that survives all sensitivity scenarios is a materially stronger case than one that wins only under the default weighting — surface this explicitly for the approval body
- Where options are defined at uneven maturity, recommend a defined further-work step to equalise definition before final selection rather than forcing a premature choice
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