Cloud FinOps Skill & MCP MCP Server

Cloud FinOps Skill & MCP

Open-source FinOps knowledge skill and MCP server for AI agents - Claude, ChatGPT, Gemini, Cursor, and any MCP-compatible client. Cloud cost optimisation across AWS, Azure, GCP and OCI, AI cost management and inference economics, Kubernetes, data platforms, allocation, chargeback, anomaly management, and named-pattern waste detection playbooks. Built by OptimNow, grounded in enterprise delivery experience.

---

Install in 5 seconds

Tool	One-step install
<img src="https://img.shields.io/badge/-Claude%20Code-D97757?logo=anthropic&logoColor=white" alt="Claude Code" height="22"/>	At the Claude Code prompt: /plugin marketplace add https://github.com/OptimNow/cloud-finops-skills.git then /plugin install cloud-finops@optimnow. The plugin is the skill only; for the six retrieval tools, add the hosted MCP connector separately (row below)
<img src="https://img.shields.io/badge/-Claude.ai%20%2F%20Desktop-D97757?logo=anthropic&logoColor=white" alt="Claude.ai / Claude Desktop" height="22"/>	Download the latest release zip, then Settings -> Skills -> Upload zip
<img src="https://img.shields.io/badge/-ChatGPT-10A37F?logo=openai&logoColor=white" alt="ChatGPT" height="22"/>	Self-host: ./install.sh --tool chatgpt --grouped (a public Cloud FinOps GPT is on the Roadmap)
<img src="https://img.shields.io/badge/-Gemini-4285F4?logo=googlegemini&logoColor=white" alt="Gemini" height="22"/>	Self-host: ./install.sh --tool gemini (a public Cloud FinOps Gem is on the Roadmap)
<img src="https://img.shields.io/badge/-Cursor-000000?logo=cursor&logoColor=white" alt="Cursor" height="22"/> <img src="https://img.shields.io/badge/-Windsurf-3DDC91?logoColor=white" alt="Windsurf" height="22"/> <img src="https://img.shields.io/badge/-Codex-412991?logo=openai&logoColor=white" alt="Codex" height="22"/> <img src="https://img.shields.io/badge/-Aider-0F172A?logoColor=white" alt="Aider" height="22"/> <img src="https://img.shields.io/badge/-Copilot-181717?logo=githubcopilot&logoColor=white" alt="Copilot" height="22"/> <img src="https://img.shields.io/badge/-Kiro%20IDE-FF6F00?logoColor=white" alt="Kiro IDE" height="22"/> <img src="https://img.shields.io/badge/-Gemini%20CLI-4285F4?logo=googlegemini&logoColor=white" alt="Gemini CLI" height="22"/>	One-liner: curl -sL https://raw.githubusercontent.com/OptimNow/cloud-finops-skills/main/install.sh | bash -s -- --tool <name>
<img src="https://img.shields.io/badge/-Auto--detect-555555?logo=gnubash&logoColor=white" alt="Auto-detect" height="22"/>	curl -sL https://raw.githubusercontent.com/OptimNow/cloud-finops-skills/main/install.sh | bash
<img src="https://img.shields.io/badge/-MCP%20hosted-7C3AED?logoColor=white" alt="MCP hosted" height="22"/>	Nothing to install: claude mcp add --transport http cloud-finops https://mcp.optimnow.io/mcp. For Claude.ai / Desktop, Settings -> Connectors -> Add custom connector with exactly that URL (exactly that form, /mcp with no trailing slash - the widget sandbox domain is derived from it)
<img src="https://img.shields.io/badge/-MCP%20package-7C3AED?logoColor=white" alt="MCP package" height="22"/>	pip install cloud-finops-mcp then add to your MCP client config (Claude Code / Cursor / Codex / Windsurf / Cline). Snippets: ./install.sh --tool mcp

Full options, troubleshooting, and the model-agnostic API loader: see INSTALLATION.md. A version-tagged zip (cloud-finops-vX.Y.Z.zip) is attached to every GitHub release.

---

What is a Skill? What is an MCP server?

A Skill is a structured knowledge folder you attach to an AI agent. Without it, general-purpose LLMs make confident but incorrect statements on FinOps topics - they miscalculate PTU break-even rates, confuse Azure and AWS reservation mechanics, and give advice that ignores how billing actually works. The answers sound plausible; they are wrong on the details that matter. The skill corrects that by injecting verified, curated FinOps knowledge directly into the model's context. The closest analogy is RAG, minus the infrastructure: no vector database, no embedding pipeline - you copy a folder and the model gains structured expertise. The same files work with Claude, GPT, Gemini, or any other model.

An MCP server (MCP: Model Context Protocol, the open standard that lets an AI client call external tools at run time) exposes the same library the other way round - as read-only tools the model queries on demand. Actual retrieval: one URL to paste, nothing to install.

Who this is for: FinOps practitioners building or evaluating AI-assisted cost analysis, cloud engineers who want a cost-aware assistant in their workflow, developers building internal FinOps agents, and Finance / IT managers evaluating the AI tooling their teams deploy. If you can copy a folder and follow the installation steps, you can use this.

Skill + installer, or MCP?

Same content, two delivery shapes - and they behave differently, because of how models use them (field-tested with the same battery of practitioner questions through both):

• Pushed into context - as a native skill (Claude Code, Claude.ai / Desktop, Kiro), or as rules files written by install.sh for tools without skill support (Cursor, Windsurf, Codex, Aider, Copilot, Gemini). The guidance is already there when the model reasons, with no per-question decision to make - which is why this surface grounds both advisory answers (commitment sizing, chargeback design, allocation methodology) and symptom questions ("my NAT gateway processes 10TB/month to S3"), where the measured probe runs show it handing over the matching runbook first try.
• Fetched on demand - the MCP server. The model must decide to call a tool per question, and that decision is this surface's real limit. Measured behaviour (August 2026 probe cycles): lookup and discovery questions ("show me the idle waste runbooks") route reliably; advisory and specific-symptom questions route since the tool descriptions carry imperative routing rules, though not on every phrasing. Its strengths: distribution (paste one URL - the right path for non-technical users and for hosts with neither skill support nor an installer target), faceted queries over the library's metadata, and interactive widgets on hosts that render MCP Apps.
• Both is legitimate. Skill loaded for the doctrine, connector added for the widgets or for hosts where the skill is not loaded. Details on the six tools: MCP server below.

What to expect in practice, on either surface. Neither the skill nor the server can see your cloud account. For "which of my X" questions ("which of my RIs are about to expire?") the deliverable is the playbook's detection query, which you run yourself - and the measured behaviour is that models tend to ask you for a data export instead of volunteering that runbook. The reliable way to get it is to ask explicitly: "check the playbook library" or "show me the runbook for this". Routing is also probabilistic, not guaranteed - the same question phrased two ways can ground differently - so when an answer arrives without a visible tool call or file read, asking for the library by name is the one-turn fix.

Using a non-Claude model through an API? Add the response contract from INSTALLATION.md ("API integration / Recommended response contract") to your system prompt so answers stay structured and billing-grounded.

Plugin and connector

The Claude plugin built from this repository contains text files only: the skill instructions in SKILL.md, the reference files and the playbooks. It runs no code, declares no MCP server and sends no data anywhere. Installing it puts FinOps knowledge into the model's context, and nothing else.

The hosted MCP connector is optional and is installed separately: as a custom connector in Claude.ai or Claude Desktop, or with claude mcp add in Claude Code (the "MCP hosted" row in the install table above). It serves the same library through six read-only tools and, on hosts that render MCP Apps, as interactive views. It is offered separately, as an MCP connector, not as part of the plugin, on purpose: the two carry the same content, and a plugin that also declared the server would load the tool definitions in every session next to the skill and put the same library into context twice. Use the plugin when you want the doctrine loaded up front, the connector when you want on-demand retrieval or the views, and both only when you want both.

---

Live prices come from OptimToken, not from this repo

This skill carries billing mechanics, which stay true for years. It deliberately does not carry current price figures, which go stale inside a packaged skill within weeks. Those live in OptimToken - LLM token rates for 250+ models and compute instance rates across seven clouds, each figure carrying its own as-of date:

<img src="https://img.shields.io/badge/-OptimToken%20web-7C3AED?logoColor=white" alt="OptimToken web" height="22"/>	optimtoken.optimnow.io - compare model and instance prices in the browser, no setup
<img src="https://img.shields.io/badge/-OptimToken%20MCP-7C3AED?logoColor=white" alt="OptimToken MCP" height="22"/>	Hosted, nothing to install. Point your client at https://ai-pricing-hub-mcp-9604f763.alpic.live/ - config snippets in INSTALLATION.md

Recommended setup on Claude: install the skill and add the OptimToken connector next to it. The skill carries the doctrine and routes pricing questions to the hub, so they get answered with a dated figure and its source rather than from a number the model remembers.

---

What this skill covers

Domain family	What is covered
AI & GenAI economics	FinOps for AI (inference economics, unit economics, ROI), agentic FinOps (agent cost anatomy, x402 / MPP), AI value management (Investment Council, stage gates), GenAI capacity planning (provisioned vs shared, spillover), self-hosted vs managed inference, open-weight vendor APIs (DeepSeek, Qwen, Kimi, GLM), AI coding tools (Cursor, Claude Code, Copilot, Windsurf, Codex)
AI platform billing	Anthropic (Fast mode, long-context cliffs, prompt caching), AWS Bedrock, Azure OpenAI Service (PTUs, spillover), GCP Vertex AI
Cloud providers	AWS (CUR / Data Exports, rightsizing, SageMaker, Savings Plans / RIs / EDP, billing hierarchy and separate invoices per business unit, pattern catalogue), Azure (Cost Management, Reservations, AHB, EA-to-MCA, pattern catalogue), GCP (CUDs, BigQuery), OCI
Data platforms	Databricks (DBU / DBCU, allocation), Microsoft Fabric (F-SKUs, CU smoothing), Snowflake (warehouses, Cortex governance)
FinOps disciplines	The FinOps Framework (22 capabilities, maturity model), tagging governance, allocation and showback (FOCUS), chargeback (Finance / tax prerequisites), anomaly management, KPIs and benchmarking, workload onboarding and M&A, Kubernetes (EKS / GKE / AKS)
SaaS & licensing	SaaS asset management (SMPs, shadow IT, renewals), ITAM collaboration (BYOL, marketplace governance, entitlements)
GreenOps	Cloud carbon measurement, carbon-aware workloads, region selection, GHG Protocol reporting
Waste detection	OptimNow's eight-category waste taxonomy, two-signal classification, three-tier confidence, WasteLine appliance for AWS - plus named-pattern runbooks across AWS, Azure, GCP and cross-cloud (full catalogue in playbooks/README.md)

The per-file catalogue with routing lives in SKILL.md - one row per reference, one row per playbook family.

Coverage, published deliberately

Coverage has two structural surfaces, answering different questions. The first is the named waste-pattern runbooks: which specific, detectable waste patterns have a ready-made playbook, per provider. A dashed cell is a known hole in the runbook catalogue, with its prioritised backlog public in docs/ROADMAP.md - it does not mean the skill cannot answer on that theme, because the reference library covers the underlying mechanics even where no runbook exists.

The second surface is the reference library mapped to the FinOps Framework: which of the 22 Framework capabilities have a reference that owns them. This is where commitment strategy, chargeback, allocation and the other advisory themes live - none of which need a runbook to be answerable.

Both maps regenerate from file frontmatter and CI fails if either drifts (details in playbook-coverage.md and fcp-coverage.md). A third, behavioural surface - does the library actually ground answers to real practitioner questions - is measured with a rotating probe battery on the maintainer side; gaps it finds land in the same public backlog.

---

Design principles

• AI cost management is a first-class domain. Most FinOps resources treat AI workloads as an edge case. This skill treats them as a primary concern, with dedicated reference files for each major AI platform.
• Visibility before optimisation. The skill follows a consistent sequence: establish what you are spending, understand what is driving it, then act. It does not recommend optimisation steps before the visibility preconditions are met.
• Provider-mechanics-first, vendor-claim-skeptical. Guidance is grounded in how billing actually works (CUR columns, Azure cost-management semantics, BigQuery export, FOCUS conformance) rather than in vendor marketing or framework positioning. Vendor sustainability and savings claims are read critically, with primary sources cited.
• Maturity is contextual, not aspirational. Verticals where cloud is not a revenue generator do not need to reach Run; Crawl plus selective Walk is the right state when cloud is a cost centre. Verticals where cloud IS the product need Run because cloud efficiency directly drives gross margin. Pushing every organisation toward the same maturity ceiling is malpractice.
• Connect cost to business value. Every recommendation answers the CFO test: what business outcome does this protect or unlock. Cost reduction without a value lens is a leak.
• Mechanics live here, price figures do not. Billing mechanics are durable; absolute prices are volatile and go stale inside a packaged file. Current-price questions route to OptimToken (see above), and any figure a reference does quote carries its date and source inline.
• FinOps is an operating discipline, not a culture. The discipline lives in allocation, anomaly management, commitment management, rightsizing, and governance, all of which produce measurable outputs. "Culture of FinOps" framing tends to substitute slideware for those outputs. In the agentic era this matters more, not less: agents execute discipline, not culture.

These principles will grow into a skills/cloud-finops/doctrine/ directory of opposable theses with their own primary sources.

---

Usage examples

These questions illustrate what the skill is designed to answer accurately - a general-purpose LLM without it produces plausible but unreliable answers to most of them, particularly on billing mechanics and capacity economics.

<div> <a href="https://www.loom.com/share/cc76d419adc64b1784e58621d6934d3e"> <p>Cloud FinOps skill - Watch Video</p> </a> <a href="https://www.loom.com/share/cc76d419adc64b1784e58621d6934d3e"> <img style="max-width:300px;" alt="Demo video: the Cloud FinOps skill answering practitioner questions in Claude" src="https://cdn.loom.com/sessions/thumbnails/cc76d419adc64b1784e58621d6934d3e-906aded8593a48f3-full-play.gif#t=0.1"> </a> </div>

• "We're spending $40K/month on AWS Bedrock and have no idea which features are driving it. Where do we start?"
• "How do I calculate the break-even utilisation rate for provisioned throughput - and should we choose Azure OpenAI PTUs or Bedrock provisioned capacity for 500K requests/day?"
• "Our monthly bill jumped from $12K to $38K after a developer enabled Fast mode in Claude Code. How do I get this under control?"
• "Should we self-host Llama 4 on rented H100s instead of paying per token - and what hidden costs do TCO calculators miss?"
• "We have $80K/month in EC2. Reserved Instances or Savings Plans - and what quick wins come first?"
• "Our client wants separate AWS invoices per business unit. Their AWS contact suggested Cost Categories - is that right?"
• "We're migrating from EA to MCA - what FinOps work do we need to do before the switch?"
• "Which VMs run for nothing, and which runbook finds them?"
• "We need to start reporting our cloud carbon emissions - where do we begin?"

---

Directory structure

cloud-finops-skills/
├── README.md                    <- This file
├── INSTALLATION.md              <- Per-tool setup, troubleshooting, API loader
├── CLAUDE.md / AGENTS.md        <- Project context for AI assistants and contributors
├── llms.txt                     <- LLM discovery index (cross-agent)
├── install.sh                   <- Cross-tool installer (12 targets)
├── mcp_server/                  <- cloud-finops-mcp PyPI package
└── skills/cloud-finops/         <- The skill - install this folder
    ├── SKILL.md                 <- Entry point + per-file routing catalogue
    ├── POWER.md                 <- Kiro IDE entry point (same references)
    ├── references/              <- The reference library, one file per domain
    └── playbooks/               <- Named-pattern runbooks (~3-8 KB each) + catalogue

---

MCP server (cross-tool, search-style retrieval)

Hosted (nothing to install) or from PyPI - both paths are in the install table above. Also listed on the MCP Registry as io.github.OptimNow/cloud-finops, and on PyPI.

Six read-only tools across two surfaces. The split is deliberate: the two content types have different shapes, and different questions attached to them.

References - the long-form provider and discipline files. Reach for these for billing mechanics, commitment strategy, allocation methodology, or any reasoning that spans patterns.

Tool	What it answers
list_references()	What guidance exists? The catalogue with its FinOps Framework facets and an approx_tokens size hint per file
get_reference(name, section?)	One guide - mechanics, decision rules, worked examples. Whole, or a single H2/H3 section when the question is narrower than the file
find_references(domain?, capability?, phase?, persona?, maturity?, persona_primary_only?)	"How should we size Savings Plans?" "What must be true before chargeback?" - routes a FinOps question to the guides that serve it (persona_primary_only cuts to the primary audience)

Playbooks - small named-pattern runbooks, one waste pattern each. Reach for these for "how do I detect and fix this specific thing".

Tool	What it answers
list_playbooks()	What cloud waste can we hunt with a ready-made runbook?
get_playbook(name)	The step-by-step runbook: symptoms, detection queries, fix, anti-pattern
find_playbooks(scope?, service?, waste_category?, confidence?)	"Which VMs run for nothing?" "Why is the NAT bill so high?" "Which of my RIs are about to expire?" - finds the runbook for a specific waste suspicion. The server cannot see your account; the runbook's detection query is the answer it hands over

Both listings carry approx_tokens per entry, because the references vary by more than tenfold - roughly 2K tokens for the smallest, over 25K for the provider pattern catalogues. The catalogues are enumerated lists, so an agent that wants one pattern family passes section (get_reference("finops-aws-patterns", section="storage")) and pays for that section instead of the whole file. Matching is case-insensitive and partial; a phrase that matches no heading returns the file's available headings rather than silently falling back to the full body.

The faceted queries are the reason this is a server and not just a folder of markdown: every file carries YAML frontmatter mapping it to a FinOps Framework capability, phase, persona and maturity gate, and a client that only fetches files cannot filter on any of it.

On hosts that support MCP Apps (SEP-1865), the tool results render as interactive widgets - a playbook explorer with facet filters and a coverage matrix, a playbook viewer with copyable detection queries and a checkable fix list, and a reference browser with a reading panel. Hosts without MCP Apps support get the plain results; nothing about the tools changes. Details in mcp_server/README.md.

---

Data handling

What the plugin runs, sends and fetches, so you can decide where it is appropriate to use it. Full policy in PRIVACY.md.

• The skill is static text. SKILL.md, the references and the playbooks are markdown files read into the model's context. They run no code, call no network endpoint and carry no credentials. The playbooks contain detection queries (billing export SQL, CLI commands) that you run in your own cloud account; nothing in this repository reads a cloud account.
• The hosted MCP connector, if you add it, sends tool calls to mcp.optimnow.io (served by the Fly.io app cloud-finops-mcp; the former cloud-finops-mcp.fly.dev host still answers but is deprecated). It is not part of the plugin (see "Plugin and connector" above). When the model calls one of the six tools, the tool arguments (a reference or playbook name, a section phrase, facet filters such as domain or scope) travel over HTTPS to that server and the matching library content comes back. The server requires no account and no authentication, holds no database and no per-user state, and serves the same public files as this repository. Its application log records facet queries that matched nothing (the filter values, never conversation text) so coverage gaps can be reviewed; the hosting platform (Fly.io, Paris region) keeps the web server's standard access log. OptimNow does not sell, share or profile from either. On hosts that render MCP Apps, the widget HTML comes from the same origin and its content security policy allows no third-party domain.
• Price lookups route to OptimToken. The skill and the server tell the model to fetch current prices from the OptimNow AI Pricing Hub (https://optimtoken.optimnow.io) instead of quoting a stale figure. That is a separate public site; whether the model opens it is its decision in the conversation.
• Nothing else. No telemetry, no analytics beacon, no update check, no package launcher, no credential read from your environment.

---

This skill is actively maintained

This is a living repository. Reference files are refreshed twice a month (around the 1st and the 15th), driven by an automated scan of around 30 data sources - cloud provider pricing pages, release notes, billing changelogs, and FinOps community publications. Changes are reviewed before being applied, so the content reflects verified updates rather than raw feed output.

AI cost management is moving particularly fast - new model releases, capacity options, and billing mechanics appear every few weeks. Watch or star this repo to be notified when updates are published.

---

Contributing

Practitioner experience is the highest-value contribution. Frameworks and vendor docs are already public; what is rare is "we tried X in production, this is what actually billed". Corrections to billing mechanics, new or improved playbooks, real-world counter-examples, and adversarial review of the recommendations are all welcome - the repo is opinionated, and it should also be falsifiable.

The full guide - contribution types, process, conventions, and what we push back on - is in CONTRIBUTING.md. One check before anything else: if your change names another OptimNow tool (an MCP tool name, an endpoint URL, a provenance field), read DEPENDENCIES.md first - most cross-repo breakage here is documentation drift that no CI check catches.

---

Adapting this skill for your organisation

Fork this repository and customise the reference files for your organisation's context: your cloud stack, your internal policies, your tag taxonomy, your preferred methodology.

A fork gives you a stable base that you can pull upstream updates into at your own pace, without overwriting your customisations. Typical customisations include:

• Adding organisation-specific tag requirements to finops-tagging.md
• Replacing generic pricing examples with your negotiated rates
• Adding reference files for internal tools or platforms not covered here
• Adjusting the methodology file to reflect your team's own approach

---

About OptimNow

OptimNow is a boutique FinOps consultancy helping organisations connect cloud and AI spend to measurable business value. Based in France with European reach.

• Website: optimnow.io
• LinkedIn: OptimNow
• GitHub: github.com/OptimNow

Open-source tools built by OptimNow:

Tool	What it does
OptimToken	Compare what 250+ models cost per request, with caching and batch factored in, plus compute instance rates across seven clouds. Also available as an MCP connector - this skill routes price questions here
AI ROI Calculator	Whether an AI project pays for itself: three-layer cost model, payback, break-even, sensitivity. Also an MCP server
AI Cost Readiness Assessment	Where your organisation stands on AI cost management
MCP for Tagging	Tag governance automation
FinOps Maturity Assessment	Crawl / Walk / Run positioning

---

Acknowledgements

This skill incorporates content derived from the following sources:

• FinOps Foundation - framework definitions, capability descriptions, and maturity model structure are based on the FinOps Framework.
• Point Five - cloud optimisation recommendations informed several provider-specific best practices and quick-win patterns.
• Tokenomics Foundation - the token complexity classes in the agentic FinOps reference are adapted from Big-T Notation by Dan Neff (Adobe), published by the Tokenomics Foundation under CC BY 4.0.

All referenced content has been adapted with additional context from OptimNow's consulting delivery experience. Any errors or opinionated interpretations are our own.

This skill is independently maintained and is not affiliated with or endorsed by the FinOps Foundation.

---

License

Licensed under CC BY-SA 4.0. See LICENSE.md.

You are free to use, adapt, and redistribute this skill - including for commercial purposes - as long as you credit OptimNow and share any derivatives under the same license.