skills.rest
matlab
matlabOFFICIAL
Explore all skills by this author →
WebsiteGitHub

matlab-discover-clusters

Discover MATLAB parallel computing clusters and manage cluster profiles.

995|122|Updated Apr 3, 2026
Search Tags:#parallel computing#hpc#matlab#mjs#Slurm#cluster discovery#cluster profiles
One-click install
npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-discover-clusters
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: matlab-discover-clusters
Source: https://github.com/matlab/matlab-agentic-toolkit/tree/main/skills-catalog/parallel-computing/matlab-discover-clusters
Command: npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-discover-clusters

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

MATLAB users often do not know which parallel computing clusters are available on their network or in the cloud, and managing cluster profiles (importing, validating, setting defaults, deleting) involves APIs that are easy to misuse or hallucinate. This Skill gives an AI agent the correct workflow and commands to discover clusters and handle the full cluster profile lifecycle.

Core Features & Use Cases

  • Cluster Discovery: Find MJS, Generic (Slurm/PBS/LSF), HPC Server, and MJSComputeCloud clusters on the network and in the cloud using the bundled discoverClusters script.
  • Profile Lifecycle Management: List, inspect, import, export, set default, validate, and delete cluster profiles using parallel.listProfiles, parallel.importProfile, parallel.validateProfile, and parallel.deleteProfile.
  • Use Case: A user receives a .mlsettings profile file from their cluster administrator and asks the agent to import it, set it as the default profile, and validate it with a capped worker count before submitting jobs.

Quick Start

Ask the agent to list your available MATLAB cluster profiles and discover any parallel computing clusters on your network.

Frequently Asked Questions about matlab-discover-clusters

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I list my MATLAB cluster profiles?▼

Use parallel.listProfiles to get all configured profile names and parallel.defaultProfile() with no arguments to read the current default. To inspect a profile, call parcluster(name) and read its Type, NumWorkers, and Host properties, wrapping the call in try/catch since MJS profiles make a network round-trip.

How do I find parallel computing clusters on my network in MATLAB?▼

Use the discoverClusters script bundled with this skill, which uses the same discovery infrastructure as the MATLAB Cluster Profile Manager UI. It returns a struct array with Type, Name, Host, NumWorkers, IsCompatible, and CorrespondingProfiles for each discovered MJS, Generic, HPCServer, or MJSComputeCloud cluster.

How do I import a .mlsettings cluster profile in MATLAB?▼

Call parallel.importProfile with the path to the .mlsettings file; it returns the imported profile's name. You can then set it as default with parallel.defaultProfile(profileName) and validate it with parallel.validateProfile before submitting jobs.

Does MATLAB cluster profile validation work in older releases?▼

The scriptable parallel.validateProfile function requires R2025a or later, and validate(cluster) on the cluster object requires R2026a. On earlier releases, use the Cluster Profile Manager UI under Home > Parallel > Create and Manage Clusters.

Why does parcluster fail with 'Unable to connect to MATLAB Job Scheduler'?▼

parcluster makes a network round-trip for MJS profiles to fetch live scheduler properties, so it throws this error when the MJS is unreachable or release-mismatched. This is expected behavior, not a code bug; wrap parcluster in try/catch and report unreachable profiles separately.

When should I not use this cluster discovery skill?▼

This skill stops at making a cluster ready to use and does not cover job submission with batch, createJob, or parfeval, nor parpool lifecycle management beyond choosing which profile to pass. GPU and parfor workflows are handled by separate skills.

Related Skills

Explore More Software Engineering Skills→
RamanEbrahimiRamanEbrahimi
3

matlab

Perform numerical computations, data analysis, and visualization with MATLAB or GNU Octave.

Community
Advanced
NVIDIA-TAONVIDIA-TAO
83

REPLACE-WITH-PLATFORM-NAME

Automate GPU job deployment on REPLACE-PLATFORM via APIs and Docker.

Official
Intermediate
AMGrobelnikAMGrobelnik
1

aii_parallel_computing

Distribute compute-intensive tasks across GPUs, CPU cores, and asynchronous I/O.

Community
Advanced
redpanda-dataredpanda-data
6

rpk

Install and configure the rpk CLI for Redpanda cluster and profile management.

Official
Advanced