Purpose
Match a capital project's specific characteristics — scope maturity, complexity, market conditions, risk appetite, and schedule pressure — to the contracting model that allocates risk to the party best able to manage it. Replaces default "lump sum by habit" decisions with a defensible, scored recommendation the project owner can take to a procurement or board approval.
When to Use
Invoke this skill when the user asks to:
- Select or recommend a contract type for a capital project
- Compare delivery models (lump sum vs. cost-plus vs. EPC vs. alliance vs. PPP)
- Decide how to allocate risk between owner and contractor before going to market
- Justify a procurement strategy to a board, sponsor, or legal review
- Reassess a contracting approach after scope, budget, or market changes
- Understand why a preferred contract model may be the wrong fit for the project
Inputs Required
Before starting, collect:
- Project description: What is being built or delivered, 2–4 sentences
- Estimated capital value: Total project budget or range
- Scope definition maturity: How well-defined the scope is (concept / FEED complete / fully designed)
- Complexity level: Technical and interface complexity (routine / moderate / high / first-of-a-kind)
- Market conditions: Contractor availability and pricing environment (buyer's market / balanced / seller's market)
- Owner risk tolerance: Appetite for cost/schedule risk (risk-averse / balanced / risk-tolerant)
- Schedule pressure: Timeline flexibility (fixed hard deadline / firm target / flexible)
- Owner capability: In-house capacity to manage design, interfaces, and change (limited / moderate / strong)
If complexity or owner capability is missing, infer from the project description and market value, and state the assumption explicitly.
Instructions
Step 1: Profile the Project Against the Five Selection Drivers
Score each driver and note its directional pull on model choice:
- Scope maturity: Low maturity → cost-reimbursable/alliance; high maturity → lump sum viable
- Complexity: High complexity → collaborative or reimbursable models; low → fixed-price efficient
- Market conditions: Seller's market inflates lump sum risk premiums → reimbursable or target-cost more competitive
- Risk tolerance: Risk-averse owner → transfer via lump sum/EPC/PPP; risk-tolerant → retain via cost-plus
- Schedule pressure: Tight schedule with immature scope → design-build or fast-track reimbursable, not fixed-price
Flag any driver conflicts (e.g., risk-averse owner + immature scope) — these are where naive model choices fail.
Step 2: Evaluate the 7 Contracting Models
For each model, assess fit against the project profile.
Model 1 — Lump Sum / Fixed Price
- Best fit: Well-defined scope, low complexity, buyer's/balanced market, risk-averse owner
- Risk allocation: Cost and productivity risk to contractor; owner retains scope-change risk
- Fails when: Scope is immature (change orders erode the fixed price and adversarial dynamics emerge)
Model 2 — Unit Rate / Measured
- Best fit: Repetitive work with quantifiable units but uncertain total quantities (earthworks, roads, drilling)
- Risk allocation: Rate risk to contractor; quantity risk to owner
- Fails when: Work is non-repetitive or interfaces dominate
Model 3 — Cost-Plus (Fixed Fee / Percentage)
- Best fit: Emergency/undefined scope, R&D, first-of-a-kind, owner with strong oversight capability
- Risk allocation: Most risk retained by owner; contractor carries little cost exposure
- Fails when: Owner lacks the capability to control cost and audit spend
Model 4 — Target Cost / GMP (Guaranteed Maximum Price)
- Best fit: Moderate scope maturity, moderate-high complexity, owner wanting shared upside
- Risk allocation: Pain/gain share around a target; GMP caps owner downside
- Fails when: Target is set before scope is stable enough to defend the baseline
Model 5 — EPC / Turnkey (Design-Build lump sum)
- Best fit: Standardized process/industrial facilities, single-point accountability desired, limited owner capability
- Risk allocation: Design, procurement, construction, and integration risk to single contractor
- Fails when: Owner wants design control or scope is not fixed at award
Model 6 — Alliance / Integrated Project Delivery
- Best fit: Very high complexity, immature scope, long duration, collaborative risk-sharing culture
- Risk allocation: Shared across an integrated commercial framework; open-book with pain/gain
- Fails when: Owner is risk-averse or lacks the maturity to run a collaborative commercial model
Model 7 — PPP / Concession (DBFO / BOOT)
- Best fit: Large public infrastructure, lifecycle operation bundled, private finance available
- Risk allocation: Design, build, finance, and operation risk transferred over concession term
- Fails when: Project is short-lived, revenue is unpredictable, or transaction costs outweigh scale
Step 3: Build a Fit-Scoring Matrix
Score each of the 7 models against the five selection drivers on a 1–5 scale (5 = strong fit). Sum to a total. Present as a table with driver columns, a total column, and a rank. Show the working — do not just declare a winner.
Step 4: Recommend the Top Model with Risk Allocation Rationale
- Name the recommended model and the runner-up
- Explain in 2–3 sentences why it wins on this project's drivers
- State the core risk allocation principle: which party carries cost risk, scope risk, schedule risk, and interface risk — and why that assignment fits capability
- Name the single biggest risk of the recommended model and one mitigation
Step 5: Flag Deal-Breakers and Conditions
List any conditions that would flip the recommendation (e.g., "If scope is not FEED-complete by tender, switch from lump sum to target cost"). Include 2–4 conditional triggers.
Step 6: Propose a Payment and Incentive Structure
For the recommended model, specify:
- Payment basis (milestone / progress / unit / cost-reimbursable + fee)
- Incentive/penalty mechanism (pain-gain %, LDs, bonus for early completion) where applicable
- Contingency and change-order approach appropriate to the model
Output Format
- Structure: Project profile summary → per-model assessment → fit-scoring matrix table → ranked recommendation → deal-breaker conditions → payment/incentive structure
- Length: 900–1,300 words
- Format: Structured document with one scoring table; per-model assessments as labeled bullets; recommendation as prose
- Sections: Selection Drivers → 7 Model Assessments → Fit-Scoring Matrix → Recommendation & Risk Allocation → Deal-Breaker Conditions → Payment & Incentive Structure
- Tone: Procurement advisor to a project board — analytical, decisive, willing to name a winner and the risk of being wrong
Quality Checklist
Before delivering output, verify:
- [ ] All 7 models are assessed for fit against the specific project — not described generically
- [ ] The fit-scoring matrix shows numeric scores across all 5 drivers for all 7 models, with a total and rank
- [ ] The recommendation names a single top model and a runner-up — no hedging across three
- [ ] Risk allocation is stated by risk type (cost, scope, schedule, interface) and tied to which party manages it best
- [ ] At least 2 deal-breaker conditions that would flip the recommendation are listed
- [ ] The payment/incentive structure matches the recommended model (e.g., no pain-gain share proposed for a pure lump sum)
Notes
- The single most common failure is choosing lump sum with immature scope — it converts every design gap into an adversarial change order; when scope maturity is low, favor target cost, alliance, or reimbursable models even if the owner prefers price certainty.
- Match risk to the party best able to control it: transferring risk a contractor cannot influence just buys an expensive risk premium.
- In a seller's market, lump sum tenders carry inflated contingency; open-book or target-cost models often deliver lower true cost even when they look less "certain" on paper.
- If the project is public and lifecycle operation matters, screen for PPP viability early — the transaction cost only pays off at scale and over a long concession.
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