Execute the build plan — introspect buildfile.m, run its dependency chain, and produce the .mltbx toolbox package. Mechanical execution with no human checkpoint. Works with any buildplan shape.
npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-build-toolbox
You execute the build pipeline. You are an operator — you introspect the plan to find the right task to invoke, run it, report results at each stage, and stop on failure. This is the only skill in the pipeline with no human checkpoint.
matlab-assess-toolbox reports READY (no blockers)buildfile.m and wants to execute itbuildfile.m exists — use matlab-create-buildfile firstmatlab-create-buildfilematlab-publish-toolbox after a successful buildmatlab-assess-toolbox| Function | Purpose |
|----------|---------|
| matlab.buildtool.Plan.load | Introspect the build plan to discover tasks and dependencies |
| buildtool | Execute the build pipeline (runs task dependency chain) |
| dir | Locate and verify the .mltbx artifact after packaging |
Before executing:
buildfile.m exists at the project rootbuildUtilities/toolboxSpecification.m and buildUtilities/tbxManifest.m exist): readiness was run, proceed with full contextbuildfile.m exists): user has their own plan not generated by matlab-create-buildfile. Skip spec/manifest checks, proceed with introspection.If buildfile.m does not exist, stop and direct the user to matlab-create-buildfile or ask them to provide one.
Load the plan and discover its structure:
plan = matlab.buildtool.Plan.load("buildfile.m");
disp({plan.Tasks.Name})
Identify the target task (in priority order):
target input if providedpackage.mltbxIdentify the dependency chain leading to the target:
% The chain executes in topological order when you run the target
% e.g., buildtool package → runs check, test, package
Report what will execute:
Build target: "package"
Dependency chain: check → test → package
Run via MATLAB, stopping on first failure:
buildtool <target>
Monitor each stage dynamically based on what's in the chain:
| Task Type | Success Condition | On Failure |
|-----------|-------------------|------------|
| CodeIssuesTask or "check" | 0 error-severity findings | Report findings, stop |
| TestTask or "test" | All tests pass | Report failures, stop |
| Coverage (if present) | Reports coverage | Log warning if below threshold (does not stop) |
| MexTask | MEX file produced | Report error, stop |
| PcodeTask | P-code files produced | Report error, stop |
| Packaging task | .mltbx file produced | Report error, stop |
Not all chains have all stages. Report only what exists in this plan.
After successful packaging, locate and verify the .mltbx. First check the task's declared Outputs (authoritative), then fall back to scanning release/ and the project root:
mltbxFile = dir(fullfile("release", "*.mltbx"));
assert(~isempty(mltbxFile), "No .mltbx file found in release/");
assert(mltbxFile(1).bytes > 0, "Package file is empty");
fprintf("Package: %s (%.1f KB)\n", fullfile(mltbxFile(1).folder, mltbxFile(1).name), mltbxFile(1).bytes / 1024);
If no .mltbx in release/, check the task's declared Outputs or scan the project root. A zero-byte file counts as a failure.
## Build Complete
- Package: release/MyToolbox.mltbx (142.3 KB)
- Target: "package"
- Chain: check → test → package
- Static analysis: 0 errors, 2 warnings
- Tests: 25/25 passed
- Coverage: 85%
- Duration: 12.3s
### Next Steps
- Run `matlab-publish-toolbox` to version-tag and release
- Test installation: matlab.addons.toolbox.installToolbox("release/MyToolbox.mltbx")
Adapt the report to what actually ran — omit sections for stages not in this plan (e.g., if no coverage task, don't report coverage).
On failure:
## Build Failed — [Stage]
### What happened
[error details]
### Suggested fix
[actionable suggestion]
### To retry
Run `matlab-build-toolbox` after fixing the issue.
.mltbx toolbox packageNo — this skill is mechanical. It executes an already-approved plan. If it fails, it reports why and stops.
.mltbx exists and has non-zero size. A zero-byte package is a silent failure./matlab-publish-toolbox — version-stamp and distribute the release----
Copyright 2026 The MathWorks, Inc.
----
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Automatically creates user-facing changelogs from git commits by analyzing commit history, categorizing changes, and transforming technical commits into clear, customer-friendly release notes. Turns hours of manual changelog writing into minutes of automated generation.
Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
React Native and Expo best practices for building performant mobile apps. Use when building React Native components, optimizing list performance, implementing animations, or working with native modules. Triggers on tasks involving React Native, Expo, mobile performance, or native platform APIs.
React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements.
Next.js best practices - file conventions, RSC boundaries, data patterns, async APIs, metadata, error handling, route handlers, image/font optimization, bundling
Use when starting feature work that needs isolation from current workspace or before executing implementation plans - creates isolated git worktrees with smart directory selection and safety verification
Take matlab/matlab-build-toolbox from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.