本文へ移動
cccskills
無料GitHub で公開

matlab-create-live-script

Create, edit, and run plain-text MATLAB live scripts (.m files) with rich text formatting, LaTeX equations, section breaks, and inline figures. Use when generating tutorials, analysis notebooks, reports, documentation, or educational content, when modifying existing live scripts, or when converting existing binary .mlx files to .m for version control. Requires R2025a+.

インストール方法を見る

含まれるファイル(3)

  • SKILL.md9.3 KB
  • manifest.yaml509 B
  • scripts/executeLiveScript.p155 B

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Live Scripts

Plain-text .m files that render as rich documents in the MATLAB Live Editor. Version-control friendly — never use binary .mlx.

When to Use

  • Tutorials, reports, analysis notebooks, or documentation
  • Interactive exploration with inline figures and equations
  • Version-controlled content (plain-text .m, not binary .mlx)
  • Converting an existing binary .mlx file to plain-text .m

When NOT to Use

  • Regular scripts without rich formatting
  • Function files
  • MATLAB older than R2025a

Converting from .mlx

To convert a binary .mlx file to a plain-text .m live script, run the following at the MATLAB Command Window. The recipe is not part of the resulting .m file:

editor = matlab.desktop.editor.openDocument(mlxPath, Visible=0);
editor.saveAs(newMPath);  % use .m extension
editor.closeNoPrompt;

Rules

  • Text lines use %[text] — NOT bare %
  • One paragraph = one %[text] line — do not hard-wrap; let the Live Editor handle line width
  • No empty %[text] lines — they render as unwanted blank space
  • Section headers: %% on its own line, then %[text] ## Title on next line
  • No blank lines in the file, except a single blank line directly before %[appendix]
  • No figure command — implicit figure creation only
  • No more than one plot per section (unless using tiled layouts)
  • No close all or clear
  • No mfilename — does not work as intended in live scripts. Hardcode filenames or use pwd.
  • Escape these characters when used as literal text (not as markdown syntax): *, _, [, ], `, \. Also escape . after a digit and # at line start.
  • Equations use single-$ inline form only: $ a = \\pi r^2 $ (no $$ ... $$). For a centered/display equation, wrap the line: %[text]{"align":"center"} $ X(k) = \\sum\_{n=0}^{N-1} x(n) e^{-j2\\pi kn/N} $
  • Inside an equation, LaTeX commands take double backslashes (\\sin, \\frac, \\pi) and markdown characters take single (\_, \*). Use a tool that takes literal text (Edit, Write, fwrite) — don't use Bash heredocs; they collapse \\ to \ even when quoted, which corrupts LaTeX equations.
  • Last list item (bulleted or numbered) ends with \
  • Every file ends with the required appendix
  • Avoid fprintf — drop the semicolon or use disp(). Output appears inline below the code that produced it, not in the Command Window.
  • Outputs should serve the reader's understanding, not verify execution

Required Appendix

Every live script must end with:

%[appendix]{"version":"1.0"}
%---
%[metadata:view]
%   data: {"layout":"inline"}
%---

Reading live scripts (Token Optimization)

When reading a live script file, ignore everything below the %[appendix] marker. The appendix contains embedded images and metadata that consume tokens without adding useful information. All code and text content appears before it.

Format Reference

SyntaxRenders as
%%Section break
%[text] # TitleH1 heading
%[text] ## SectionH2 heading
%[text] **bold**Bold
%[text] *italic*Italic
%[text] `code`Monospace
%[text] <u>text</u>Underlined text
%[text] $ a = \\pi r^2 $Inline equation
%[text]{"align":"center"} $ ... $Centered/display equation
%[text] - itemBullet
%[text] - last \Last bullet
%[text] 1. itemNumbered list
%[text] 2. last \Last numbered item
%[text] [text](url)External hyperlink
%[text] [text](internal:id)Internal link to an anchor
%[text] %[text:anchor:id] ...Anchor (link target)
%[text:tableOfContents]{"heading":"..."}Table of Contents

IDs (anchors, and any other id-bearing element): letters, digits, and underscores only. No hyphens — my-section won't bind; use my_section. For anchors, place the marker immediately after %[text] at the start of the line.

Tables

%[text:table]
%[text] | Method | Result |
%[text] | --- | --- |
%[text] | Trapezoidal | 1.9998 |
%[text:table]

Example

%[text] # Sinusoidal Signals
%[text] Examples of sinusoidal signals in MATLAB.
%[text:tableOfContents]{"heading":"Contents"}
%[text] - sine waves
%[text] - cosine waves \
x = linspace(0,8*pi);
%%
%[text] ## Sine Wave
plot(x,sin(x))
title('Sine Wave')
xlabel('x (radians)')
ylabel('sin(x)')
grid on
%%
%[text] ## Cosine Wave
plot(x,cos(x))
title('Cosine Wave')
xlabel('x (radians)')
ylabel('cos(x)')
grid on
%%
%[text] ## Summary
%[text] The sine and cosine functions are $ \\pi/2 $ radians out of phase.

%[appendix]{"version":"1.0"}
%---
%[metadata:view]
%   data: {"layout":"inline"}
%---

Common Patterns

Mathematical Explanations with Equations

%[text] ## Theory
%[text] The discrete Fourier transform is defined as:
%[text]{"align":"center"} $ X(k) = \\sum\_{n=0}^{N-1} x(n) e^{-j2\\pi kn/N} $
%[text] where $ x(n) $ are the time-domain samples and $ k $ indexes the frequency bins.

Code with Inline Comments

%%
%[text] ## Data Processing
%[text] Load and filter the data, then visualize the results.
data = load('measurements.mat');
filtered = lowpass(data, 0.5);  % Apply lowpass filter
plot(filtered)
title('Filtered Data')

Tiled Layouts for Comparison

Use only when side-by-side comparison is important to the illustration:

%%
%[text] ## Comparison of Methods
tiledlayout(1,2)
nexttile
plot(method1)
title('Method 1')
nexttile
plot(method2)
title('Method 2')

Workflow

  1. Plan — Title, setup, analysis sections, summary
  2. Write — %[text] for text, %% for sections, appendix at end. Use a tool that takes literal text (Edit, Write, fwrite). Don't use Bash heredocs (cat > file <<EOF); they collapse \\ to \ even when quoted, which corrupts LaTeX equations.
  3. Validate (if MATLAB attached) — Run the code through evaluate_matlab_code to confirm it executes cleanly. Note the wall-clock time — Embed Outputs takes about the same.
  4. Embed Outputs (optional, if MATLAB attached) — Run executeLiveScript("<absolute-path>.m") to save each section's outputs (plots, displayed values) inline next to the code that produced them.
  5. Verify Rendering (optional, if MATLAB attached) — Run export("<absolute-path>.m", "<absolute-path>.html") to produce an HTML approximation of the rendered document. Read it back to confirm equations are typeset, figures appear inline, %[text] directives don't leak as literal text, etc. Delete the .html when done — it's a verification artifact, not a deliverable.

The Write step is the load-bearing one — Write alone produces a valid live script the user can open and run. Validate, Embed Outputs, and Verify Rendering are progressive enhancements that require an attached MATLAB session. Without one, stop after Write. After Embed Outputs, the file is rewritten to disk by MATLAB — re-read before any further edits.

executeLiveScript

Bundled in the scripts/ folder of this skill. Add to path before first use:

addpath(fullfile(skillRoot, "scripts"));   % skillRoot = directory containing this SKILL.md

Calling executeLiveScript(filePath) returns nothing on success. Runtime errors inside the script do not raise exceptions — the script writes them into the appendix as "dataType":"error" blocks; grep the saved .m to find them.

Run Validate before Embed Outputs. If a cell errors during Embed Outputs, that cell becomes an error block and outputs in cells after it are stripped from the file. Run through evaluate_matlab_code first.

If Embed Outputs errors with "Nested Live Editor execution" or hangs past the Validate time: retry once with the MATLAB desktop visible. If that also fails, the on-disk file from Write is still valid — skip Embed Outputs and let the user run the script themselves.

Checklist

Before finishing a live script, verify:

  • File has .m extension
  • Sections use %% alone on its own line, followed by %[text] ##
  • No blank lines or empty %[text] lines (except one blank line directly before %[appendix])
  • Each paragraph is a single %[text] line (no hard-wrapping)
  • One plot per section (unless tiled layout)
  • Bulleted and numbered lists end with backslash on last item
  • LaTeX command backslashes are doubled in the saved file: \\sin, \\frac, \\pi
  • No figure commands
  • No close all or clear at start
  • No mfilename
  • Appendix is present and correctly formatted
  • Outputs serve the reader, not the developer

Not Yet Supported

Minor features planned for a future revision:

  • Interactive controls (sliders, dropdowns, numeric inputs)
  • Hide Code View (output view that hides the source code)

Copyright 2026 The MathWorks, Inc.


レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Guide for accessing financial and economic data in MATLAB using the Datafeed Toolbox. Covers Bloomberg (market data via bloomberg/blp/bloombergHypermedia), FRED (Federal Reserve economic data via fredrs), Haver Analytics (economic data via haver/haverdirect/haverview), and LSEG Datastream (historical data via datastreamws). Use when connecting to any of these data providers from MATLAB.

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

Read 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.

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

Analyze 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.

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

Design and analyze electrically large antenna structures using MATLAB Antenna Toolbox. Covers reflector antennas (parabolic, Cassegrain, Gregorian, offset, corner, cylindrical, spherical, custom STL), reflectarrays and reconfigurable intelligent surfaces (RIS), antennas installed on platforms (vehicles, aircraft, ships, satellites), and radar cross section (RCS) analysis. Includes solver selection (MoM-PO, PO, MoM, FMM), mesh control, and GPU acceleration. Use when the user wants to design a dish/reflector antenna, reflectarray, analyze an antenna on a platform, or compute RCS.

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

Analyze 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, row 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.

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

S-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).

日本語の概要は準備中です。原文の説明を表示しています。

matlab/matlab-agentic-toolkit1,1492026年10月9日 更新

matlab のスキルをすべて見る

このスキルの問題を報告する