Manages MATLAB projects for Simulink workflows: path management, file registration, labels, source control configuration, and project lifecycle. Use when creating projects, adding models/dictionaries/requirements to projects, configuring labels for automation, fixing broken model references, or setting up source control for Simulink artifacts.
How this skill is triggered — by the user, by Claude, or both
Slash command
/model-based-design-core:managing-simulink-projectsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
MATLAB projects coordinate Simulink workflows — they manage the MATLAB path (which Simulink uses to resolve model references, data dictionaries, and requirements), track file membership for source control, and provide label-based automation for selecting targets in CI/CD pipelines.
MATLAB projects coordinate Simulink workflows — they manage the MATLAB path (which Simulink uses to resolve model references, data dictionaries, and requirements), track file membership for source control, and provide label-based automation for selecting targets in CI/CD pipelines.
building-simulink-modelstesting-simulink-modelsgenerate-requirement-draftssimulating-simulink-modelsFour principles that prevent the most common failures:
Registration ≠ path. addFile registers a file with the project for source control tracking. addPath puts a folder on the MATLAB path so Simulink can resolve artifacts by filename. These are independent operations — you almost always need both.
Filename-only resolution. Data dictionaries (set_param(...,'DataDictionary',...), addDataSource), model references, and requirements all resolve via MATLAB path using filename only — never full paths. The containing folder must be on the project path.
addPath is not recursive. Each subfolder containing resolvable artifacts needs its own addPath call.
Single-project constraint. Only one project can be open at a time. Creating or opening a new project silently closes the current one (running its shutdown scripts). currentProject throws an error when no project is open — it does not return empty.
Always:
addPath(proj, folder) for every folder containing dictionaries, referenced models, requirements, or profilesaddDataSource, set_param(...,'DataDictionary',...), and slreq.createLinkSimulinkCacheFolder and SimulinkCodeGenFolder: proj.SimulinkCacheFolder = fullfile(proj.RootFolder, 'work')updateDependencies(proj) before querying proj.Dependencies.graphml extension for DependencyCacheFile.gitignore for excluding derived files from source control.slmx link-store files with the project (guarded by isfile check)Ask First:
removeFile then disk delete)Never:
proj.IgnoredFilePatterns — throws "Feature not supported" for all project types from createProjectcat.Labels — the correct property is cat.LabelDefinitions'string' as a label DataType — valid types are 'none', 'char', 'double', 'integer', 'logical'findFile (singular) for label queries in R2024a+ — use findFiles (plural). ONLY use findFile with R2023a/bclear all in startup scripts — it wipes the proj variable| Intent | Reference |
|---|---|
| Add files/folders to project, manage project path | references/path-and-file-management.md |
| Set up labels for automation pipelines | references/labels-and-automation.md |
| Link data dictionaries to models | references/data-dictionaries.md |
| Configure model references across folders | references/model-references.md |
Configure source control, cache folders, .gitignore | references/source-control-and-caching.md |
| Diagnose broken references or health check failures | Start with Verification below, then route to the relevant reference |
Run after every project modification:
results = runChecks(proj);
updateDependencies(proj);
deps = proj.Dependencies; % digraph
If runChecks reports issues, inspect which folders are missing from the project path — this is the root cause of most failures.
Copyright 2026 The MathWorks, Inc.
npx claudepluginhub matlab/simulink-agentic-toolkit --plugin model-based-system-engineeringCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.