matlab/matlab-agentic-toolkit

Aerospace Fundamentals performs aerospace unit, time, coordinate-frame, and rotation conversions using Aerospace Toolbox. Use it when converting between frames, Julian dates, or rotation forms such as Euler angles, DCMs, quaternions, and Rodrigues vectors.

Build RoadRunner Scenario From OpenSCENARIO recreates a RoadRunner scenario from an OpenSCENARIO 1.x file using MATLAB. Use it to convert a .xosc into a programmatic scenario script, including positions, trajectories, references, and logic that can be mapped.

Compute Aerospace Environment calculates atmosphere, gravity, wind, magnetic field, geoid, ephemeris, space weather, and Earth orientation properties for aerospace analysis and navigation corrections. Use it when MATLAB prompts need standard atmosphere, gravity, magnetic, or IERS/Aerospace Toolbox environment data.

Deploy Embedded Code configures MATLAB Coder and Embedded Coder for embedded hardware targets, including memory limits, hardware settings, and PIL/SIL verification. Use it when generating code for microcontrollers or boards like STM32 and Raspberry Pi.

Driving Data Importer loads raw driving sensor logs into scenariobuilder objects and runs preprocessing such as sync, crop, offset correction, and timestamp normalization. Use it when converting datasets like KITTI, nuScenes, Waymo, ROS bags, or CSV/MAT/video files for later scenario work.

Embedded AI Deployment helps deploy neural networks to embedded hardware with MATLAB, Simulink, and Embedded Coder. Use it to import, compress, simulate, and generate C or CUDA for Cortex-M/A/R, x86, or GPU targets.

Euro Ncap Variants generates Euro NCAP seed scenarios, variants, descriptors, and reports with the ADT support package. Use it to translate between drivingScenario and RoadRunner or compute NCAP scores.

Import Deep Learning Models brings PyTorch, ONNX, and Keras/TensorFlow models into MATLAB as dlnetwork objects. Use it when converting .pt2, .pt, .onnx, or .keras files and when checking import errors or output validation.

Interactive Curve Fitting guides no-code curve and surface fitting in the Curve Fitter app. Use it to explore fit types, exclude outliers, compare fits, and export results to MATLAB code, the workspace, figures, or Simulink lookup tables.

Lanelet2 To RRHD Converter converts Lanelet2 .osm maps into RoadRunner HD Map .rrhd files in MATLAB. Use it when importing Lanelet2 road networks into RoadRunner and preserving lanes, junctions, signs, markings, and barriers.

MATLAB AI Model Deployment generates MEX, C/C++, or CUDA code from PyTorch ExportedProgram or LiteRT models for use in applications or Simulink. Use it when you need to verify inference in MATLAB and then deploy the model as a library, executable, or GPU-accelerated build.

MATLAB Classify Tabular Data compares classifiers on tabular datasets, including data preparation, cross-validation or holdout evaluation, hyperparameter optimization, and statistical tests. Use it to find strong models and judge whether accuracy differences are significant.

MATLAB Generate Code generates, verifies, and refines C/C++ or CUDA code from MATLAB using MATLAB Coder, Embedded Coder, GPU Coder, and MATLAB Test. It also helps write codegen-ready MATLAB with %#codegen, coder.* directives, and code generation constraints.

MATLAB Share Content helps you upload MATLAB files to GitHub, MATLAB Drive, or File Exchange and generate Open in MATLAB Online links. Use it when you want to share code or open local files in MATLAB Online.

MATLAB Train Network trains, evaluates, and exports neural networks in MATLAB using trainnet, dlnetwork, and related APIs. Use it to migrate legacy deep learning code, run inference, or speed up custom training and dlaccelerate workflows.

Matlab Create Experiment plans an Experiment Manager experiment from code or a problem description. It is used to sweep parameters, compare configurations, and generate the experiment setup or training function.

Matlab Machine Learning Apps opens Classification Learner and Regression Learner to train, compare, evaluate, and export tabular machine learning models. Use it when you want to work in the app, test on held-out data, inspect plots, or export a trained model.

Matlab Sumo Co-Simulation builds Simulink models that connect to Eclipse SUMO for synchronized traffic co-simulation. Use it to read vehicle state, control an ego vehicle, generate traffic, and validate ADAS scenarios through TraCI.

RRHD Authoring builds RoadRunner HD Map entities in MATLAB using the roadrunnerHDMap API. Use it to create lanes, boundaries, junctions, signs, signals, barriers, and parking from scratch or from converted map data.

RoadRunner Asset Mapping provides lookup tables for RoadRunner project asset paths used when converting map formats to RRHD. It helps resolve lane markings, signs, signals, barriers, and static object assets from source map data.

RoadRunner Core sets up the MATLAB RoadRunner connection and manages project, scene, and scenario lifecycle tasks. Use it to initialize rrApp, set world origin or scene bounds, check status, and close RoadRunner safely.

RoadRunner Scenario Authoring programmatically creates RoadRunner scenarios from MATLAB using roadrunnerAPI. Use it to add actors, routes, phase logic, and scenario patterns like cut-ins or pedestrian crossings.

RoadRunner Scenario Simulation runs RoadRunner scenarios from MATLAB and Simulink, including stepping, actor control, observers, sensors, logs, and scenario variables. Use it when you need to simulate or co-simulate an existing scenario, not to author scenes or scenarios.

RoadRunner Scene Import loads RRHD and OpenDRIVE files into a RoadRunner scene for visualization and verification. Use it when importing driving maps, checking conversion results, or configuring scene build options in MATLAB.

Scenario Builder generates simulation-ready driving scenes, trajectories, road surfaces, and 3D assets from recorded GPS, camera, lidar, and actor-track data. Use it to build or export scenarios for RoadRunner, drivingScenario, OpenSCENARIO, OpenDRIVE, OpenCRG, or Unreal Engine.