Execution and Simulation Engine for Trading Strategies R package tradesimr version 0.18.7 An R-native trading simulation package with a C++ execution core that turns strategy intentions and explicit orders into simulated trades, positions, cash, profit and loss, risk, and performance outputs under configurable execution, margin, funding, and cost assumptions. The package provides historical replay, incremental exchange stepping, durable event tables, append-only agent command logs, registered assets, per-agent shared-cash cross-margin live accounts, AI agent competitors, scheduled live-feed stepping, strategy-backed AI agents with diagnostics, calibrated and coordinated multi-asset market simulation with static covariance, AR-GARCH, factor, and regime models, durable per-feed simulation state, profile-aware heterogeneous inventory and margin execution with atomic mixed-profile order groups, optional portfolio-margin enforcement through a multi-asset C++ step kernel, local live-service APIs, import/export helpers, separate replay, live-state, and agent dashboard exports, and installed local orchestration scripts. It is designed to consume signals, order intents, or target exposure decisions from compatible strategy packages and market data from compatible adapters. Software Depends: R (>= 4.2.0) Imports: data.table, Rcpp, R6 LinkingTo: Rcpp Suggests: testthat, fst, ggplot2, jsonlite, lubridate, plumber, strategyr, zoo Oliver Zhou <oliver.yxzhou@gmail.com> Comprehensive R Archive Network (CRAN) Oliver Zhou [aut, cre] MIT + file LICENSE (https://CRAN.R-project.org/package=tradesimr/LICENSE) 2026-10-08 application/tgz https://CRAN.R-project.org/package=tradesimr doi:10.32614/CRAN.package.tradesimr