Practical Model Context Protocol (MCP) resources for Cloud FinOps: pricing, budgets, anomaly checks, automation — with security guardrails.
- ✅ Download a client
- ✅ Run your first MCP
- ✅ Explore the FinOps MCPs and pick the one that best fits your use case
- /foundations → background notes, whitepapers, blog summaries
- /servers → registry of available MCP servers (pricing, tagging, governance)
- /clients → tested MCP clients (Claude, Cursor, VS Code, Q, etc.)
- /tutorials → runnable guides (step by step)
- /finops-use-cases → applied scenarios (budgeting, anomaly detection, tagging compliance)
- /tooling-governance → security checklists, threat models, deployment guidance
- /presentations → slides, abstracts, LinkedIn drafts
- /resources → external links (FinOps WG docs, repos, talks, videos)
MCP is a protocol that lets LLMs act as agents by safely connecting to external tools (servers) like AWS Cost Explorer, a GCP BigQuery dataset with billing exports, an Azure storage account holding cost data, or 3rd-party cloud finops solutions like Vantage.
In FinOps, MCP unlocks:
- Faster cost simulations
- Real-time tagging compliance
- Forecasting and Cost Simulations
- Cost Optimization recommendations
But also raises governance and security challenges — this repo addresses both sides.
We welcome contributions:
- Fork the repo and create a branch
- Add your MCP server, tutorial, or use case
- Open a PR with a clear description
See CONTRIBUTING.md for details.
New to MCP or FinOps? Start with issues labeled good first issue.
- Join discussions in the FinOps Foundation Slack
- Follow updates on LinkedIn
- Share your use cases, raise issues, propose servers
This project is licensed under the Apache 2.0 License.