matlab-create-buildfile

$npx mdskill add matlab/matlab-agentic-toolkit/matlab-create-buildfile

Generate a MATLAB buildfile.m for automated build, test, and package tasks.

  • Creates a buildfile.m to automate static analysis, testing, coverage, and packaging.
  • Depends on MATLAB's matlab.buildtool framework and existing project structure.
  • Scans project for source folders, test folders, and existing build artifacts.
  • Outputs a buildfile.m file with customizable tasks and thresholds.

SKILL.md

.github/skills/matlab-create-buildfileView on GitHub ↗
---
name: matlab-create-buildfile
description: "Generate a MATLAB buildfile.m with tasks for static analysis, testing, coverage reporting, and packaging. Use after matlab-create-project when the project structure is in place and you need repeatable build automation."
license: MathWorks BSD-3-Clause
metadata:
  author: MathWorks
  version: "1.0"
---

# matlab-create-buildfile — Build Plan Generator

You generate a `buildfile.m` that defines the repeatable build/test/package pipeline using MATLAB's `matlab.buildtool` framework.

## When to Use

- After `matlab-create-project` has set up the project structure
- User says "set up the build" or "create a buildfile"
- Project has code and tests but no build automation

## When NOT to Use

- A `buildfile.m` already exists and works — use `matlab-build-toolbox` to execute it
- User wants to run the build, not create it — use `matlab-build-toolbox`
- No MATLAB project exists yet — use `matlab-create-project` first

## Inputs

- **project_root**: Path to the project (default: current directory)
- **coverage_threshold** (optional): Line-coverage percentage to warn below (default: 80)
- **warning_threshold** (optional): Max warnings before check fails (default: 0 = strict)

## Workflow

### Step 1 — Assess What Exists

Scan the project for:
- Source folder — one of (in priority order):
  1. `toolbox/` — the standard toolbox-design-guidelines layout (everything that ships)
  2. `+packagename/` — namespace-package layout (from matlab-create-project)
  3. `source/` or `src/` — generic source folder
- `tests/` — test files to run
- MEX source files — C/C++/Fortran files (`.c`, `.cpp`, `.cxx`, `.F`, `.f90`) in folders like `mex/`, `src/mex/`, `c_src/`, or at the project root. Presence indicates the project needs a `MexTask`.
- `toolboxPackaging.prj` — packaging configuration (produced by Toolbox Packaging Tool)
- Existing `buildfile.m` — update rather than replace
- `buildUtilities/toolboxSpecification.m` — interface spec (for context on what the toolbox exposes)

Record the detected structure — the generated buildfile must reference actual paths.

### Step 2 — Generate `buildfile.m`

Use built-in task types (`CodeIssuesTask`, `CleanTask`, `TestTask`, `MexTask`) where they exist, and custom function-based tasks only where built-in tasks lack needed behavior (coverage reporting, packaging).

**Task strategy:**
- **`clean`** — built-in `CleanTask`
- **`check`** — built-in `CodeIssuesTask` (SARIF output, threshold enforcement)
- **`mex`** — built-in `MexTask` (only if MEX source files detected in Step 1). Use `MexTask.forEachFile` when multiple MEX sources exist. Output folder is `toolbox/` (or source folder) so MEX files ship with the toolbox.
- **`test`** — built-in `TestTask` with `.addCodeCoverage()`. Produces JUnit XML test results AND a `.mat` coverage file for programmatic inspection by the coverage task. The built-in task supports incremental builds — it skips when source/tests are unchanged.
- **`coverage`** — custom function-based task that loads the `.mat` coverage results from the test task, logs per-file coverage, and warns if below the threshold. It does NOT fail the build — coverage is advisory, not a gate.
- **`package`** — custom function-based task (no built-in equivalent for toolbox packaging).

**Include comments in the generated buildfile** that explain design choices — particularly why a task is custom vs. built-in, what tradeoffs that creates, and how the user could switch approaches.

Use `scripts/buildfile-template.m` as the base template. Apply these adaptation rules:
- Replace `"toolbox"` with the actual source folder detected in Step 1
- Replace `"tests"` if tests live elsewhere
- Replace `0.80` with the user's coverage threshold (as a decimal)
- Replace `0` in `WarningThreshold` with the user's warning threshold
- If MEX source files were detected, add a `MexTask` with appropriate source paths and output folder. Set `plan("test").Dependencies` to include `"mex"` so tests run after MEX compilation.
- If no MEX source files exist, omit the `mex` task entirely (don't generate dead code).
- If no `toolboxPackaging.prj` exists, use the programmatic variant from `references/buildfile-variants.md`
- Set `plan("package").Outputs` to match the actual output path

### Step 3 — Present the Plan

```
## Build Plan — [Toolbox Name]

| Task | Type | Description | Dependencies | Fail condition |
|------|------|-------------|--------------|----------------|
| clean | CleanTask | Remove derived artifacts | — | — |
| check | CodeIssuesTask | Static analysis (SARIF output) | — | Any error; any warning (strict) |
| mex | MexTask | Compile MEX files (if detected) | — | MEX compilation fails |
| test | TestTask | Run tests + produce coverage | check, mex (if present) | Any test failure |
| coverage | Custom | Report coverage, warn if below threshold | test | — (advisory only) |
| package | Custom | Build .mltbx from toolboxPackaging.prj | coverage | Package file not produced |

Default: `buildtool` → runs check + test + coverage
Full pipeline: `buildtool package` → check → [mex] → test → coverage → package
List tasks: `buildtool -tasks`
CI invocation: `matlab -batch "buildtool check test coverage package"`

### Artifacts Produced

| File | Format | Consumer |
|------|--------|----------|
| results/code-issues.sarif | SARIF v2.1.0 | GitHub Code Scanning, VS Code |
| results/test-results.xml | JUnit XML | CI test reporting |
| results/coverage.xml | Cobertura XML | CI coverage tools |
| results/coverage.mat | MAT-file | Coverage report task (programmatic) |
| release/My_Toolbox.mltbx | Toolbox installer | End users |

How would you like to proceed?
> A) **Approve** — write the buildfile as shown
> B) **Adjust** — modify tasks, thresholds, or dependencies
> C) **Skip** — don't create a buildfile now
```

### Step 4 — Persist

**If `buildfile.m` does NOT exist:** Write it to the project root. Add `results/` and `release/` to `.gitignore` if it exists.

**If `buildfile.m` already exists:** Do NOT edit it directly. Instead:
1. **Read the existing test task** to determine where coverage data is produced (path and format). The existing test task may write Cobertura XML, `.mat`, or both — and may use a different output directory (e.g., `reports/` vs. `results/`). The coverage report task MUST reference the actual output path and format produced by the test task.
2. Show a diff or code block of the proposed additions/modifications (new tasks, updated dependencies, new local functions).
3. Explain what each change does and why.
4. **Wait for explicit user approval** ("yes", "go ahead", "looks good") before applying any edits.
5. Only after the user confirms, apply the changes to the existing `buildfile.m`.

This approval gate prevents surprising edits to working build automation that the user may have customized.

## Output

- `buildfile.m` — the complete build plan

## Checkpoint

**Yes** — user reviews the task chain before it's written. They can adjust order, thresholds, and which tasks are included.

## Key Rules

- **Comment design decisions in the generated code.** Every task should have a comment explaining whether it's built-in or custom and WHY. For custom tasks, explain what the built-in alternative lacks and what tradeoff the custom approach introduces. Include a commented-out snippet showing how to switch to the simpler alternative. The buildfile is a teaching artifact — the user must be able to understand and maintain it without re-running this skill.
- **Use built-in tasks where they exist.** `CodeIssuesTask`, `CleanTask`, `TestTask`, and `MexTask` are battle-tested — don't reimplement them as function tasks.
- **TestTask handles testing AND coverage production.** Use the built-in `TestTask` with `.addCodeCoverage()` to produce both Cobertura XML (for CI) and `.mat` (for programmatic threshold checking). This gives incremental build support — the task skips when source/tests are unchanged.
- **Coverage reporting is a separate custom task.** The `coverageTask` loads coverage data, logs per-file results, and warns if below threshold — but does NOT fail the build. Coverage is advisory. To make it a hard gate, the user can replace the warning `context.log` with `context.assertTrue`.
- **Coverage task must match actual test output.** When adding a coverage task to an existing buildfile, read the test task (or its helper) to determine the actual coverage output path and format. If the test task produces Cobertura XML (e.g., `reports/codecoverage.xml`), parse the `line-rate` attribute from the XML root. If it produces `.mat` (from `TestTask.addCodeCoverage`), use `coverageSummary`. Never hardcode `results/coverage.mat` without verifying that the test task actually writes it.
- **MexTask for MEX compilation.** When MEX source files are detected, use the built-in `MexTask` (or `MexTask.forEachFile` for multiple sources). Place output in the source/toolbox folder so compiled MEX files ship with the toolbox. Tests must depend on the mex task.
- **Custom tasks use `context`.** Always accept the `context` argument and use `context.log()` for output, `context.assertTrue()` for failure conditions. NEVER use `disp()`, `fprintf()`, or `warning()` for status output in task functions — always `context.log()`. NEVER use bare `assert()` for failures — always `context.assertTrue()`.
- **Single test run.** The built-in `TestTask` with `.addCodeCoverage()` instruments coverage in the same run that checks pass/fail — never run tests twice.
- **Package from PRJ.** Load `ToolboxOptions` from `toolboxPackaging.prj` — this is the single source of truth for toolbox identity, files, and metadata. Only fall back to programmatic construction if no PRJ exists.
- **Never hardcode the version in packageTask.** The version must be read from `buildUtilities/toolboxSpecification.m` (if it exists) or from the PRJ file — never written as a literal string in `buildfile.m`. Hardcoded versions create drift: `matlab-publish-toolbox` updates `toolboxSpecification.m` before packaging, but a hardcoded `opts.ToolboxVersion = "1.0.0"` silently overrides it. The spec is the single source of truth for version.
- **Output to `release/`.** The `.mltbx` goes in `release/` (not source-controlled). Replace spaces with underscores in the filename for cross-platform compatibility.
- **Produce CI artifacts.** Always emit SARIF (code issues), JUnit XML (test results), Cobertura XML (coverage), and `.mat` (for coverage reporting) — these are the standard formats consumed by GitHub Actions, Azure DevOps, Jenkins, and the coverage task.
- **Declare outputs on package task.** Setting `.Outputs` lets `CleanTask` know what to delete and enables incremental build support.
- **`DefaultTasks = ["check" "test" "coverage"]`.** Running bare `buildtool` should validate code quality including coverage. Packaging is an explicit action (`buildtool package`).
- **Update, don't replace.** If `buildfile.m` already exists, add missing tasks rather than overwriting existing customization. Always propose changes as a plan and wait for user approval before editing.
- **Detect structure, don't assume.** The source folder varies (`toolbox/`, `+pkg/`, `source/`). Always verify what exists before generating.
- **Omit MEX task if no MEX sources.** Don't generate a mex task with placeholder paths — only include it when C/C++/Fortran source files are actually detected.

## Next Steps

- `/matlab-assess-toolbox` — validate readiness across all checks before building
- `/matlab-build-toolbox` — execute the build plan and produce the `.mltbx` artifact

----

Copyright 2026 The MathWorks, Inc.

----

More from matlab/matlab-agentic-toolkit

SkillDescription
matlab-access-datafeed>
matlab-add-awgnRead BEFORE writing any code that adds Additive White Gaussian Noise (AWGN) to signals and converts between SNR, Eb/No, Es/No, and per-subcarrier SNR for communications simulations, using awgn(), convertSNR(), berawgn(). The default MATLAB patterns for AWGN (e.g., 'measured' option, manual SNR formulas) produce subtly incorrect results. This skill specifies the correct calling conventions, required function usage, and critical anti-patterns that must be avoided.
matlab-analyze-ams-waveformAnalyze AMS waveform data using Mixed-Signal Blockset utilities: phase noise measurement, clock jitter, anti-aliased resampling, timing measurements, lock time, INL/DNL, ADC/DAC calibration, HSpice import. Use when analyzing time-domain voltage from PLL/VCO/clock simulations, measuring phase noise from variable-step solver output, computing jitter, or resampling non-uniform data.
matlab-analyze-dataAnalyze data using MATLAB. Use when the task involves tables, timetables, time-series data, numeric arrays, sensor matrices, or gridded data — including but not limited to exploring, filtering, sorting, cleaning, transforming, aggregating, smoothing, padding, trimming, and answering questions about data. MATLAB provides extensive, easy-to-use built-in functions for these workflows with no additional products required.
matlab-analyze-dependenciesAnalyze the effective toolbox file set to produce a Dependency Manifest — classify all transitive dependencies as included, product, add-on, or external-unresolved, then present resolution options with tradeoffs. Use after matlab-define-toolbox-api when the spec is approved.
matlab-analyze-emS-parameters, insertion loss, fields, currents, mesh control, and solver selection for RF PCB performance validation. TRIGGER: user asks to compute S-parameters, analyze insertion/return loss, extract fields or currents, compare MoM vs FEM, or control mesh for any RF PCB component. Invoke BEFORE writing sparameters() or solver code — API is non-obvious. SKIP: designing or creating components (use the specific matlab-design-pcb-* skill), material/stackup setup only (use matlab-manage-pcb-material), optimization sweeps (use matlab-optimize-pcb-design), PDN/IR-drop analysis (use matlab-analyze-pcb-pdn).
matlab-analyze-installed-antennaAnalyze antennas installed on electrically large conducting platforms using MATLAB Antenna Toolbox. Loads platform geometry from STL/STEP/IGES, installs antenna elements, selects electromagnetic solvers (MoM-PO, FMM, MoM), and computes patterns, impedance, coupling, and efficiency. Use when the user wants to model an antenna on a vehicle, aircraft, ship, satellite, or other large structure.
matlab-analyze-pcb-pdnPDN DC voltage/current analysis, IR drop, design rule checking, and multi-net batch analysis on imported PCB layouts. TRIGGER: user asks about power integrity, PDN analysis, IR drop, voltage distribution, current density, power nets, or design rule checking on a PCB. Invoke BEFORE writing code — the PDN API chain is specialized and non-obvious. SKIP: importing a PCB file (use matlab-read-pcb-layout), EM field/S-parameter extraction (use matlab-analyze-em), material/stackup setup only (use matlab-manage-pcb-material), transmission line design (use matlab-design-pcb-txline).
matlab-analyze-rcsCalculate and visualize monostatic and bistatic radar cross section (RCS) using MATLAB Antenna Toolbox. Computes RCS of platforms, antennas, and arrays with PO, MoM, and FMM solvers, supporting HH/VV/HV/VH polarization, GPU acceleration, and near-field observation. Use when the user wants to compute, plot, or analyze radar cross section.
matlab-analyze-rf-propagationAnalyze RF propagation and plan wireless sites using MATLAB Antenna Toolbox. Creates transmitter/receiver sites, computes signal strength, coverage maps, SINR, line-of-sight, and ray tracing in geographic or indoor environments. Supports multiple propagation models (free-space, close-in, Longley-Rice, ray tracing, rain/gas/fog), custom terrain, building data, and directional antennas. Use when the user wants to compute coverage, signal strength, path loss, SINR, ray tracing, or plan a wireless network.