OPEN SOURCE
Framework4D and FinOps run on two deterministic analysis engines, published as standalone, MIT-licensed projects anyone can install, audit, and run without CloudSealed.
github.com/cloudsealedCloud billing cost anomaly detection
Detects structural waste in cloud billing exports (AWS/GCP/Azure) using a rolling-median + MAD baseline, resistant to the masking effect that breaks textbook mean+standard-deviation detectors. Python library, CLI, HTTP service, and GitHub Action.
pip install cloudsealed-jitView on GitHub Deterministic architecture risk scoring
Scores single point of failure, excessive coupling, and scalability gap from a declared system inventory — every score ships with a rule-by-rule breakdown, not a black box. C#/.NET 10, NuGet package, CLI, and GitHub Action.
dotnet add package CloudSealed.ML.CoreView on GitHub MCP server for AI agents
Gives AI agents (Claude Code, Claude Desktop, Cursor) direct access to both engines as tools, so they call a real deterministic calculation instead of guessing. Independent from CloudSealed's own hosted MCP server for account data — see Developers → MCP for that.
uvx cloudsealed-mcpView on GitHub A finding you can't verify isn't a finding you can act on. Publishing the underlying engines — methodology, benchmarks, and scoring rules included — means every result CloudSealed produces can be independently checked, not taken on faith.
Source code, issues, and contribution guides live on GitHub.
github.com/cloudsealed