Hpc Gmsh logo

Hpc Gmsh

Organization
SciMate-AI
hpc-gmsh

Build, review, debug, and automate Gmsh geometry and meshing workflows. Use when working with `.geo` scripts, the Gmsh Python API, GEO versus OpenCASCADE modeling, physical groups, mesh-size fields, transfinite or recombine options, boundary layers, structured/unstructured algorithm selection, mesh partitioning, periodic meshes, high-order elements, mesh export and solver handoff (FEniCS, OpenFOAM, Elmer, SU2, etc.), CAD/STL import, or Gmsh CLI and meshing failures.

Overview

PublisherSciMate-AI
RepositoryHPC-Skills
Skill namehpc-gmsh
Stars
97
Forks
10
Bundled files
19
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 19 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by SciMate-AI on GitHub. Read the source before you install it.

Installation

Install the Hpc Gmsh AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/SciMate-AI/HPC-Skills/tree/main/skills/hpc-gmsh
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/SciMate-AI/HPC-Skills.git /tmp/HPC-Skills
mkdir -p .claude/skills
cp -r /tmp/HPC-Skills/skills/hpc-gmsh .claude/skills/hpc-gmsh
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Hpc Gmsh in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Hpc Gmsh on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Hpc Gmsh is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

HPC Gmsh

Treat Gmsh as a geometry-plus-meshing stack. Choose the modeling kernel first, then make topology, sizing, and structured-meshing decisions explicit before exporting downstream.

Start

  1. Read references/modeling-kernels-and-geometry.md before creating or repairing a Gmsh model — covers GEO versus OpenCASCADE selection, boolean operations, extrusions, transformations, CAD/STL import, embedded and compound entities, synchronization rules.
  2. Read references/physical-groups-and-entity-tagging.md when solver handoff depends on boundary or region tags — covers export behavior, physical group stability after booleans, downstream solver requirements matrix.
  3. Read references/algorithm-selection-guide.md when choosing 2D or 3D meshing algorithms — covers MeshAdapt, Delaunay, Frontal-Delaunay, HXT, parallel meshing, optimization, element order, subdivision.
  4. Read references/mesh-size-and-field-control.md when choosing mesh sizes — covers the full size determination hierarchy, all field types (Distance, Threshold, Box, Ball, Cylinder, BoundaryLayer, MathEval, AutomaticMeshSizeField, etc.), combining fields, size callbacks.
  5. Read references/boundary-layer-and-refinement-recipes.md when boundary layers, distance-based refinement, or CFD-specific mesh recipes are needed — covers BL parameters, y+ estimation, fan points, wake refinement, complete 2D CFD recipe.
  6. Read references/transfinite-and-recombine-playbook.md when the mesh should be structured, swept, or recombined — covers TransfiniteCurve/Surface/Volume syntax with grading, Recombine, structured extrusion with Layers, periodic meshes, high-order, complete hex workflow.
  7. Read references/cli-and-python-api-playbook.md when choosing between .geo, CLI, and the Python API — covers complete CLI flag reference, Python API patterns for geometry, meshing, fields, partitioning, periodic meshes, CAD import.
  8. Read references/import-export-and-solver-handoff.md when importing CAD/STL or exporting meshes — covers all output formats, MSH4.1/MSH2 format details, element type table, downstream handoff recipes for FEniCS, OpenFOAM, Elmer, SU2, ABAQUS, Code_Aster, mesh partitioning.
  9. Read references/error-recovery.md when meshing or geometry construction fails — covers diagnostic commands, recovery decision tree, geometry/meshing failure recovery sequences, reporting template.
  10. Read references/error-pattern-dictionary.md for fast matching of specific Gmsh failure signatures — 14 patterns covering wrong kernel, missing physical groups, transfinite incompatibility, synchronize missing, boolean tag instability, 3D boundary recovery, negative Jacobians, periodic mismatch, and more.

Scenario Recipes

Load the relevant recipe when the task involves:

  • references/algorithm-selection-guide.md — 2D/3D algorithm selection matrices, parallel meshing, optimization passes, element order, subdivision for all-quad/all-hex
  • references/boundary-layer-and-refinement-recipes.md — BoundaryLayer field setup, y+ targeting, fan points, distance+threshold refinement, box/ball/cylinder refinement, MathEval sizing, complete 2D CFD airfoil recipe, 3D BL strategies
  • references/transfinite-and-recombine-playbook.md — structured hex/quad meshing, TransfiniteCurve grading (Progression, Bump), periodic mesh setup (affine transforms), extruded mesh with Layers, high-order elements
  • references/import-export-and-solver-handoff.md — STEP/IGES/BREP/STL import and healing, MSH4.1/MSH2/VTK/SU2/CGNS/MED format details, downstream solver physical group requirements, mesh partitioning with Metis

Work Sequence

  1. Choose one modeling path and stay consistent:
    • GEO kernel for simple scripted primitives, explicit point/curve control, and hybrid discrete models
    • OpenCASCADE for CAD-style solids, booleans (fuse/cut/fragment), fillets, pipes, lofts, and STEP/IGES import
  2. Build clean topology before tuning mesh size:
    • create all geometry
    • perform all boolean operations
    • call synchronize()
    • inspect entities and boundaries
  3. Define physical groups after geometry is finalized:
    • physical groups encode solver-facing semantics (inlet, outlet, walls, fluid, solid)
    • use getBoundary() and getEntitiesInBoundingBox() to identify surfaces after booleans
    • if physical groups are defined, only tagged elements are exported (use Mesh.SaveAll = 1 to override)
  4. Select the meshing algorithm:
    • 2D default: Delaunay (5) for speed, Frontal-Delaunay (6) for quality, MeshAdapt (1) for robustness
    • 3D default: Delaunay (1) for robustness, HXT (10) for parallel speed
    • embedded entities and size fields require Delaunay or HXT
  5. Configure mesh sizing:
    • start coarse and uniform, then add local refinement one field at a time
    • disable MeshSizeFromPoints, MeshSizeFromCurvature, MeshSizeExtendFromBoundary when fields are authoritative
    • combine multiple fields with Min and set as Background Field
  6. Use structured meshing only where topology supports it:
    • transfinite requires 3- or 4-sided surfaces with compatible node counts on opposite edges
    • extrusion with Layers produces structured hex/prism
    • recombine converts tris to quads (use Blossom algorithm + smoothing for quality)
  7. Optimize and validate before export:
    • optimize tets: gmsh.model.mesh.optimize("Netgen")
    • high-order: setOrder(2) then optimize("HighOrder")
    • check quality: getElementQualities(tags, "minSICN")
  8. Export in the format matching the downstream solver:
    • verify physical groups are present in the output
    • verify element types and mesh dimension match solver expectations

Guardrails

  • Do not mix GEO kernel and OpenCASCADE in the same model.
  • Do not define physical groups before synchronize() — entity tags may be stale.
  • Do not rely on raw entity tags staying stable after boolean operations — use BooleanFragments + the output map.
  • Do not force transfinite or recombine settings onto incompatible topology.
  • Do not use Mesh.SaveAll = 1 with MSH2 if physical group info is needed — MSH2 discards it.
  • Do not treat export format or mesh version as an afterthought — the receiving solver decides what metadata matters.
  • Do not add multiple competing size fields without combining them through a Min field.
  • Do not use embedded entities with algorithms other than Delaunay or HXT — they will be silently ignored.
  • Do not skip optimize("HighOrder") after setOrder(2) — curved elements on complex geometry will have negative Jacobians.
  • Do not mesh 3D without first verifying the 2D surface mesh is valid — 3D boundary recovery depends on it.

Additional References

Load these on demand:

  • references/modeling-kernels-and-geometry.md — GEO versus OpenCASCADE, boolean operations, extrusions, transformations, CAD/STL import, embedded/compound entities
  • references/physical-groups-and-entity-tagging.md — physical group syntax, export behavior, downstream solver requirements
  • references/algorithm-selection-guide.md — 2D and 3D algorithm matrices, parallel meshing, optimization, element order
  • references/mesh-size-and-field-control.md — complete field reference (Distance, Threshold, Box, Ball, Cylinder, BoundaryLayer, MathEval, AutomaticMeshSizeField), size hierarchy, callbacks
  • references/boundary-layer-and-refinement-recipes.md — BL setup, y+ targeting, wake refinement, complete CFD recipes
  • references/transfinite-and-recombine-playbook.md — structured meshing, periodic meshes, high-order, hex workflow
  • references/cli-and-python-api-playbook.md — all CLI flags, Python API patterns, .geo scripting reference
  • references/import-export-and-solver-handoff.md — format details, element types, solver handoff recipes, partitioning
  • references/error-pattern-dictionary.md — 14 failure patterns with symptoms, diagnosis, and fixes
  • references/error-recovery.md — diagnostic commands, recovery decision tree, reporting template

Reusable Templates

Use assets/templates/ when a concrete starting point is faster than building from scratch:

  • minimal_rectangle_2d.geo — simplest OCC rectangle with physical groups
  • occ_box_python.py — 3D box with boundary extraction via Python API
  • parametric_cylinder_in_box.geo — parametric external flow geometry with distance-based refinement
  • airfoil_bl_2d.py — 2D airfoil with boundary layer, wake refinement, and fan points
  • periodic_channel_hex.py — structured hex channel with periodic BCs (DNS/LES)
  • step_import_mesh.py — STEP import, healing, physical groups, parallel meshing
  • gmsh-batch-mesh.sh — minimal batch meshing script
  • gmsh-mesh-slurm.sh — SLURM-submitted batch meshing

Outputs

Summarize:

  • chosen modeling kernel and why
  • physical groups and their intended downstream meaning
  • meshing algorithm (2D and 3D) and why
  • mesh-size strategy: fields used, MeshSizeMin/Max, background field
  • structured-meshing or recombine choices if any
  • element order and optimization passes
  • export format and the exact failure class if the workflow is being repaired
  • mesh statistics: element counts by type, minimum quality metric

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Hpc Gmsh AI skill do?

Build, review, debug, and automate Gmsh geometry and meshing workflows. Use when working with `.geo` scripts, the Gmsh Python API, GEO versus OpenCASCADE modeling, physical groups, mesh-size fields, transfinite or recombine options, boundary layers, structured/unstructured algorithm selection, mesh partitioning, periodic meshes, high-order elements, mesh export and solver handoff (FEniCS, OpenFOAM, Elmer, SU2, etc.), CAD/STL import, or Gmsh CLI and meshing failures.

Why use Hpc Gmsh on TypingMind?

Because you install it once and use it with any model. Hpc Gmsh is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Hpc Gmsh in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/SciMate-AI/HPC-Skills/tree/main/skills/hpc-gmsh. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Hpc Gmsh?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Hpc Gmsh?

As many as you like. As long as a model supports skills, you can use Hpc Gmsh with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Hpc Gmsh AI skill free?

Yes. It is published on GitHub by SciMate-AI under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇