Understand the system. Then make it yours.
Short explanations first, exact details when you need them. These guides describe the current PragmaChange v1 implementation.
Get the management stack running and follow the path to a first release.
02 / GUIDE ↗How it worksSee how request data becomes local decisions in Nginx.
03 / GUIDE ↗Partition keysUnderstand separate client histories, hashing, and warm-up.
04 / GUIDE ↗Pipelines & featuresDefine extractors, windows, and measurements for one endpoint.
05 / GUIDE ↗Datasets & trainingImport representative data and train a matching model.
06 / GUIDE ↗Decisions & thresholdsRead scores and choose allow, check, and block behavior.
07 / GUIDE ↗Deployment & rollbackLoad a release into Nginx and switch back safely.
08 / GUIDE ↗TroubleshootingWork through common data, model, and runtime issues.
09 / GUIDE ↗Current limitsKnow exactly what v1 does and does not observe.
10 / GUIDE ↗API referenceDownload the v1 OpenAPI specification.
11 / GUIDE ↗Reference clientsAdapt examples for Node.js, Python, C#/.NET, and Rust.
12 / GUIDE ↗Offline recordingsCreate and upload datasets with or without a client library.
13 / GUIDE ↗Application behaviorModel named actions and workflows in a separate service.
14 / GUIDE ↗Remote serversInspect proxies, deploy reviewed releases, and verify loaded state.
15 / GUIDE ↗Assisted setupReview configuration drafts and prepare training.
16 / GUIDE ↗Native developmentRun the individual services directly on Linux.
Reading path
New here? Start with the quickstart, then read partition keys and decisions. If you are deploying, keep the Nginx guide open.
Source: PragmaChange and reference clients. Repository access is currently restricted. The setup guides use the product checkout; this website remains a separate static project.