One number out of ten, built from five dimensions with published weights. Here is what each one measures, how we get the number today, and what is changing.
REST or GraphQL completeness, auth standards, versioning.
Weight 25%Workflow triggers, bulk operations, no-code support.
Weight 25%Native AI, copilots, AI-powered suggestions.
Weight 20%Real-time events, reliability, event coverage.
Weight 15%API docs quality, SDKs, and working examples.
Weight 15%Agentic score = 25% API quality + 25% automation + 20% AI features + 15% webhooks + 15% documentation. Each dimension is scored from 0 to 10. A stack score averages the tools in it and highlights the weakest dimension.
Today, scores are estimated from public information: product pages, API references, and documentation. We are building measured scoring that reads API docs, webhook catalogues, SDKs, and changelog activity directly. Each listing says whether its score is estimated or measured.
Listings start from public directories and product pages, then vendors can correct facts with a link we can verify. We never invent customers, ratings, or review counts.
CloudCare is owned by the same group as Humonix (an AI consultancy) and Allometry (an AI platform). When their products or partners appear in CloudCare, they are scored with the same published method, labelled as related, and never ranked higher because of who owns us. Vendors cannot pay to change a score.