# CloudSealed

> CloudSealed is an autonomous cloud governance and forensic audit platform (Framework4D + FinOps).
> It audits architecture risk, security posture, cost waste, and compliance readiness across multi-cloud environments.
> Findings are powered by deterministic, auditable analysis engines — not black-box ML — ensuring every diagnostic score traces back to explicit rules, mathematical models, and verifiable rationale.

Full documentation index and technical specifications for LLMs and AI Agents: [https://cloudsealed.com/llms-full.txt](https://cloudsealed.com/llms-full.txt)

## Core Capabilities & Services

- [Cloud Forensic Audit](https://cloudsealed.com/start-audit): Multi-dimensional forensic assessment across security, cost, reliability, and regulatory compliance.
- [Framework4D Governance](https://cloudsealed.com/): Comprehensive audit across four critical vectors:
  1. **Security & Posture**: Identity & access auditing, misconfigurations, lateral movement risks, and automated attack-surface mapping.
  2. **FinOps & Cost Efficiency**: Anomaly detection, idle/orphaned asset elimination, rightsizing, and reservation commitment optimization.
  3. **Architecture & Reliability**: Single points of failure (SPOF) detection, blast-radius containment, coupling analysis, and scalability ceilings.
  4. **Compliance & Governance**: Automated mapping against SOC 2 Type II, ISO 27001, CIS Benchmarks, NIST CSF, and AWS Well-Architected Framework.
- [FinOps Cloud](https://cloudsealed.com/finops-cloud): Continuous cloud spend intelligence, unit economics tracking, and automated budget anomaly alerts.
- [Cloud Security Scanner](https://cloudsealed.com/scanner): Rapid non-invasive audit scans identifying immediate exposure risks and cost leaks.
- [Open Source Engines](https://cloudsealed.com/open-source): Publicly verifiable, MIT-licensed core analytical engines for developers and security teams.

## Supported Cloud Stack & Ecosystem

- **Cloud Providers**: Amazon Web Services (AWS), Google Cloud Platform (GCP), Microsoft Azure, Oracle Cloud Infrastructure (OCI), Hetzner Cloud, DigitalOcean, and Hybrid Cloud deployments.
- **Infrastructure & Containers**: Kubernetes (K8s), Docker, Terraform, OpenTofu, AWS CloudFormation, Pulumi.
- **Observability & Telemetry**: Prometheus, Grafana, Datadog, New Relic, Splunk, AWS CloudWatch, GCP Cloud Logging, Azure Monitor.
- **Security & Compliance Baselines**: CIS Benchmarks v8, SOC 2 Type II, ISO/IEC 27001, NIST CSF 2.0, PCI-DSS v4.0, HIPAA Security Rule, AWS Well-Architected Framework.

## Diagnostic Models & Methodology

- **FinOps Anomaly Detection (Deterministic MAD)**:
  - Uses Rolling-Median + Median Absolute Deviation (MAD) rather than Gaussian mean-plus-standard-deviation.
  - Immune to the masking effect where extreme cost spikes distort historical baselines.
  - Open Source: [JIT-Optimization-Engine](https://github.com/cloudsealed/JIT-Optimization-Engine) (`pip install cloudsealed-jit`).
- **Architecture Risk Scoring (Graph & Invariant Analysis)**:
  - Deterministic evaluation of system inventories evaluating SPOFs, circular dependencies, excessive service coupling, and scaling limitations.
  - Every finding provides an exact mathematical score and transparent breakdown of violated architecture rules.
  - Open Source: [Predictive-ML-Core](https://github.com/cloudsealed/Predictive-ML-Core) (`dotnet add package CloudSealed.ML.Core`).
- **AI Agent Integration (Model Context Protocol)**:
  - Hosted MCP Server: `https://mcp.cloudsealed.com` (authenticated account access for enterprise audit querying).
  - Open Source MCP Server: [cloudsealed-mcp](https://github.com/cloudsealed/cloudsealed-mcp) (`uvx cloudsealed-mcp`) allowing AI assistants (Claude Code, Cursor, Codex) to run audits locally against cloud manifests.

## Developer & API Resources

- [Developer Portal](https://cloudsealed.com/developers): Overview of developer tools, authentication, and integrations.
- [REST API Reference](https://cloudsealed.com/developers/api): Programmatic access to audit runs, findings, risk metrics, and report exports.
- [Model Context Protocol (MCP)](https://cloudsealed.com/developers/mcp): Native tool integration for Claude Desktop, Claude Code, Cursor, and custom agentic workflows.
- [Claude Skill](https://cloudsealed.com/developers/skill): Pre-built prompt workflows and skills for Anthropic Claude.
- [SDKs](https://cloudsealed.com/developers/sdks): Official client libraries for TypeScript, Python, Go, and C# (.NET 10).
- [Webhooks](https://cloudsealed.com/developers/webhooks): Real-time event notifications for high-severity security findings and budget breaches.

## Company & Legal

- [About CloudSealed](https://cloudsealed.com/about)
- [Company Information](https://cloudsealed.com/company-info)
- [Schedule an Audit](https://cloudsealed.com/schedule)
- [Contact Team](https://cloudsealed.com/contact)
- [Privacy Policy](https://cloudsealed.com/privacy)
- [Terms of Service](https://cloudsealed.com/terms)