Purpose
Produce an evidence-based maturity assessment of an organization's cloud financial management practices — scored per FinOps Framework domain and capability — and translate the gaps into a prioritized, sequenced set of improvement actions. Replaces vague "we need to be better at cloud cost" conversations with a tier-by-tier scorecard and a concrete next-move plan.
When to Use
Invoke this skill when the user asks to:
- Assess how mature their FinOps or cloud cost management practice is
- Benchmark cost governance against the FinOps Framework
- Build a business case for investing in FinOps tooling, headcount, or processes
- Identify the highest-leverage gaps in cost allocation, chargeback, anomaly detection, or commitment management
- Plan the next phase of a cloud cost program after early wins have plateaued
- Prepare a maturity scorecard for CFO, VP Eng, or board-level reporting
Inputs Required
Before starting, collect:
- Annual cloud spend: Approximate total across all providers (e.g., $2M/yr on AWS + Azure)
- Cloud providers: Which platforms are in use (AWS, Azure, GCP, multi-cloud)
- Team structure: Who owns cost today — dedicated FinOps team, finance, engineering, or nobody
- Current tooling: Native cost tools (Cost Explorer, Azure Cost Management) or third-party (Cloudability, CloudHealth, Vantage, etc.)
- Cost allocation state: Tagging coverage, account/project structure, how spend maps to teams
- Existing practices: What they already do — showback/chargeback, budgets, RIs/Savings Plans, anomaly alerts
- Primary driver: What triggered this assessment — runaway costs, upcoming audit, growth, board pressure
If team structure or tooling is unspecified, assume a decentralized model with native cloud tools only and note the assumption.
Instructions
Step 1: Establish the Maturity Scale
Use the FinOps Foundation Crawl / Walk / Run model consistently across every capability:
- Crawl: Ad hoc, reactive, manual. Little or no automation. Coverage under ~50%. Processes are inconsistent or undocumented.
- Walk: Defined processes exist and are followed by some teams. Partial automation. Coverage 50–90%. Metrics tracked but not universally acted on.
- Run: Automated, proactive, and embedded in daily engineering and finance workflows. Coverage >90%. Continuous improvement with clear ownership and KPIs.
Assign a tier to each capability below, then roll capabilities up into a phase-level tier.
Step 2: Assess the Inform Phase (Visibility & Allocation)
Score these capabilities:
- Cost Allocation: Tagging/labeling strategy, account/project hierarchy, % of spend that maps cleanly to a team, product, or cost center. Untagged spend above 10% caps this at Walk.
- Showback / Chargeback: Is spend reported back to teams (showback) or billed to their budgets (chargeback)? Frequency, granularity, and whether teams actually see it.
- Reporting & Analytics: Dashboards, cadence, unit economics (cost per customer, per transaction, per feature), and who consumes them.
For each: state current tier, cite the evidence from inputs, and name the single blocker preventing the next tier.
Step 3: Assess the Optimize Phase (Efficiency & Rate)
Score these capabilities:
- Anomaly Detection: Automated alerting on unexpected spend spikes, time-to-detection, false-positive handling, and who responds.
- Commitment Management: Reserved Instances, Savings Plans, Committed Use Discounts — coverage %, utilization %, and whether commitments are actively managed vs. set-and-forget.
- Workload / Rate Optimization: Rightsizing, autoscaling, storage tiering, spot/preemptible usage, idle resource cleanup. Whether optimization is a recurring practice or a one-time cleanup.
For each: state tier, cite evidence, name the blocker. Flag any capability where quick, high-ROI wins exist (e.g., low commitment coverage on stable workloads).
Step 4: Assess the Operate Phase (Governance & Culture)
Score these capabilities:
- Governance & Policy: Budgets, guardrails, provisioning policies, approval workflows, and enforcement mechanisms.
- Team Enablement & Accountability: Do engineers see cost impact of their decisions? Is cost a KPI? Is there shared language between finance and engineering?
- Continuous Improvement / FinOps Operating Model: Is there a defined FinOps function, executive sponsorship, and a feedback loop that turns Inform + Optimize outputs into policy?
For each: state tier, cite evidence, name the blocker.
Step 5: Build the Maturity Scorecard
Compile all nine capabilities into a table: Capability | Phase | Current Tier | Target Tier | Primary Blocker. Assign a phase-level tier (the lowest capability tier in a phase generally caps the phase, since gaps compound).
Step 6: Prioritize Improvement Actions
Rank 8–10 improvement actions by expected impact ÷ effort. For each action include:
- The capability it advances and the tier jump it enables (e.g., Crawl → Walk)
- Expected financial or operational impact (quantified where spend data allows — e.g., "increasing Savings Plan coverage from 20% to 60% on stable compute ≈ 15–25% savings on that portion")
- Effort level (Low / Medium / High) and rough owner
- Dependencies — what must exist first (allocation before chargeback; visibility before anomaly detection)
Sequence Inform-phase gaps first when foundational visibility is missing — optimization and governance cannot mature on top of poor allocation.
Step 7: Deliver the 90-Day Roadmap
Group prioritized actions into three 30-day waves:
- Days 1–30: Foundational fixes — usually tagging, allocation, and baseline reporting
- Days 31–60: First optimization and detection wins
- Days 61–90: Governance, accountability, and operating-model steps State the target maturity tier per phase at the end of 90 days.
Output Format
- Structure: Executive summary → per-phase assessments (Inform, Optimize, Operate) → scorecard table → prioritized action list → 90-day roadmap
- Length: 1,000–1,400 words
- Format: Structured document with one summary scorecard table and a numbered priority action list; supporting analysis in short prose + bullets
- Sections: Executive Summary (overall maturity + top 3 gaps) → Inform Phase → Optimize Phase → Operate Phase → Maturity Scorecard → Prioritized Actions → 90-Day Roadmap
- Tone: FinOps practitioner advising a CFO — analytical, direct, quantified where possible, no cloud-vendor cheerleading
Quality Checklist
Before delivering output, verify:
- [ ] All nine capabilities across the three phases are assigned a Crawl/Walk/Run tier
- [ ] Each tier is justified with specific evidence from the inputs — not assigned arbitrarily
- [ ] The scorecard table includes current tier, target tier, and primary blocker for every capability
- [ ] Prioritized actions are ranked by impact-to-effort and include dependencies — not listed alphabetically
- [ ] At least one action includes a quantified savings or impact estimate tied to the stated cloud spend
- [ ] The 90-day roadmap sequences foundational visibility work before optimization and governance
Notes
- The most common failure pattern is over-investing in optimization tooling while allocation is still at Crawl — you cannot optimize what you cannot see or attribute.
- Untagged/unallocated spend is the single most diagnostic metric; if it's above 20%, nearly every downstream capability is capped at Crawl regardless of tooling.
- Chargeback maturity should not be pushed faster than the organization's cultural readiness — premature chargeback without accurate allocation destroys trust between finance and engineering.
- If spend data is thin, keep impact estimates as percentage ranges rather than fabricating dollar figures.
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