Multidimensional Computer Adaptive Testing with the Shadow Testing routine
-
Updated
Nov 20, 2017 - R
Multidimensional Computer Adaptive Testing with the Shadow Testing routine
Proksi is a set of tools and applications that help you shadow test your web application.
A Request Mirroring and Shadow Testing module for Caddy 2
Darkest is a shadow testing and dark launching library for Java.
A pure-Python shadow deployment and canary release router for ML models — failure-isolated shadow calls, deterministic hash-based traffic splitting, and statistical promotion gates in one library.
Go gateway that re-encodes JSON tool results (TOON, toonx) to cut LLM input tokens, and switches each route on only after a shadow test proves answers stay as good as JSON and the net saving is worth it.
echo middleware that logs requests and responses for shadow testing purposes.
Shadow-test TypeSafe Jev (or any decision model) against the LLM classifier it would replace, then pick thresholds from real traffic.
Monitor an AI assistant in production and test the monitoring itself: tracing, online evals, drift (PSI), p95 latency, guardrails, shadow releases, incident triage and feedback loops, with CI gates and negative controls. No API key needed.
Mirror production traffic to a shadow deployment and diff the responses structurally, with declarative suppression of timestamps, IDs, ordering and float precision.
To associate your repository with the shadow-testing topic, visit your repo's landing page and select "manage topics."