Skip to main content

MCP server for ELF600 electromagnetic field analysis: 31 curated topics (incl. motor workflows and ELF MAGIC -> Cauer Ladder Network synthesis) + public workflow recipes + 486 ELF-runnable public .mai/.meg sample decks for PM/SPM/SRM/reluctance/hysteresis motors, induction motors, WPT, MRI, IH, transformers, and accelerator electromagnets + Python-interface team28 seed manifest + 1141 help files + 332 examples + 146 wiki pages + Python ctypes API — 24 tools, all searchable.

Project description

ELF-mcp-server

PyPI License: BSD-3-Clause Python: 3.10+

MCP server providing ELF600 electromagnetic field analysis documentation — file formats, solver options, element types, workflow recipes, and ELF-runnable public input decks for the ELF600 BEM-based electromagnetic analysis suite (MAGIC magnetostatic, ELFIN electrostatic, BEAM particle tracking).

This server does not execute ELF600 simulations — it provides curated documentation and public .mai/.meg input decks that AI coding assistants (Claude Code, Cursor, etc.) can consult while authoring ELF input files.


Features

24 tools + 1 prompt providing curated docs, workflow recipes, ELF-runnable public sample decks, and raw access to ELF600 help HTM, example inputs, vendor wiki, and Python ctypes API:

Tool family Purpose Files
elf_usage(topic) 31 curated topics — high-level recipes (knowledge.py)
elf_help_*(...) Help HTM files from C:/ELF600/help/ 1141 files, 1.18M chars
elf_examples_*(...) Example .mai/.mei/.txt plus 100-card playbook from C:/ELF600/examples/ 332 files, 533k chars
elf_sample_decks_*(...) Lab-authored ELF-runnable public .mai/.meg sample decks 486 cases, 972 input files
elf_recipe_*(...) Workflow decision cards for elements, PRE/SOL blocks, outputs, checks, and pitfalls public-safe recipes
elf_wiki_*(...) Vendor wiki pages from elf.co.jp PukiWiki 146 pages, 211k chars
elf_python_*(...) Python ctypes API + configs from C:/ELF600/bin/ 15 files, 246k chars

Each _* family has 3 tools: _index, _search(query), _get(path). The examples family also has elf_examples_playbook(limit=100), which summarizes 100 .mai examples as compact cards with detected SOL blocks, element families, feature tags, companion .mei/.model files, and reuse hints. The recipe family also has elf_plan_workflow(goal), which chooses a short public-safe recipe sequence from a natural-language analysis goal. The sample deck family has elf_sample_decks_index/search/get/playbook for ELF-runnable public .mai/.meg decks. The Python family also has elf_python_team28(): a compact 28-case seed manifest from the public motor cases for ELF Python-interface orchestration. team28 is not a normal ELF GUI/CLI deck-execution workflow. Solver outputs, comparison metrics, executable orchestration state, and Python-interface runtime state are not bundled.

MCP quick start

For MCP clients, start with elf_overview() to discover the server surface and public boundary. The most useful calls while authoring ELF/MAGIC inputs are:

  • elf_sample_decks_playbook(limit=20, family="pm_square") for compact cards over the public PM motor decks
  • elf_sample_decks_playbook(limit=20, family="spm") for surface-PM motor decks with stator coils, rotor/stator iron, and pickup coils
  • elf_sample_decks_playbook(limit=20, family="srm") for switched-reluctance motor decks with salient iron and phase-pair excitation
  • elf_sample_decks_playbook(limit=20, family="induction_cage_10") for induction motor cage decks with transient eddy-current pickup patterns
  • elf_sample_decks_playbook(limit=20, family="application") for transformer MRI gradient-coil, WPT coupled-coil, induction-heating, and accelerator electromagnet application decks
  • elf_sample_decks_playbook(limit=20, query="Loop10") for the 10-cycle learning-loop decks across WPT, MRI, SR motor, SPM, IH, reluctance motor, hysteresis motor, transformer, and accelerator electromagnet families
  • elf_sample_decks_search("HBCN FLUM", ext="mai") to find reusable input patterns
  • elf_sample_decks_search("SPM HBRM FLUM", ext="mai") to find surface-PM motor setup patterns
  • elf_sample_decks_search("SRM reluctance FLUM", ext="mai") to find switched-reluctance motor setup patterns
  • elf_sample_decks_search("WPT MOMC FLUM", ext="mai") to find wireless power-transfer AC coupling patterns
  • elf_sample_decks_search("MRI OHM2 FREQ", ext="mai") to find AC shielding and eddy-current setup patterns
  • elf_sample_decks_search("induction motor cage OHM2 FLUM", ext="mai") to find induction-motor cage and pickup setup patterns
  • elf_sample_decks_search("accelerator electromagnet FLUM", ext="mai") to find coil/yoke electromagnet setup patterns
  • elf_sample_decks_search("IH induction-heating MOMC", ext="mai") to find induction-heating AC conductor setup patterns
  • elf_sample_decks_get("motor/pm_cosine_pickup_72/pm001/pm001.mai") to open a concrete public deck
  • elf_python_team28() to inspect the Python-interface seed manifest

This MCP server is documentation and input-deck retrieval only; it does not launch ELF, run solvers, manage licenses, or publish validation outputs.

ELF/MAGIC application input authoring

ELF/MAGIC is useful for magnetostatic and AC magnetic input authoring when the model is expressed as .mai analysis control plus .meg mesh data. This server turns that knowledge into MCP tools:

  • 402 public motor input-deck pairs covering 2-pole, 4-pole, 6-pole, 8-pole, cosine-remanence PM pickup families, 10 explicit SPM motor examples, and 10 SRM switched-reluctance examples, 10 induction cage examples, plus loop-reviewed SPM, SR motor, synchronous-reluctance motor, and hysteresis motor families
  • 84 public application input-deck pairs covering transformer core/pickup coupling, MRI gradient-coil/eddy-current shield patterns, WPT coupled coils, IH induction-heating workpieces, and accelerator electromagnets
  • playbook cards that expose each deck's SOL blocks, PRE keywords, element families, feature tags, and reuse hints
  • curated motor topics for air-gap field, flux linkage/back-EMF pickup, polarity/angle conventions, force outputs, and eddy-current setup
  • Python-interface team28 seed manifest for higher-level orchestration, without bundling runtime state or solver outputs

Useful entry points are elf_usage(topic="ipm_motor"), elf_usage(topic="motor_radia_bridge"), elf_recipe_search("motor pickup"), and elf_sample_decks_playbook(limit=50, family="pm_square"), elf_sample_decks_playbook(family="spm"), or elf_sample_decks_playbook(family="srm"), or elf_sample_decks_playbook(family="induction_cage_10"). For non-motor applications, start with elf_sample_decks_playbook(family="application"), elf_sample_decks_search("transformer FLUM"), elf_sample_decks_search("WPT MOMC FLUM"), or elf_sample_decks_search("MRI OHM2 FREQ"), elf_sample_decks_search("IH induction-heating MOMC"), or elf_sample_decks_search("accelerator electromagnet FLUM").

Public lint

Before publishing, run:

python -m pytest
elf-mcp-server --selftest
elf-mcp-policy-lint

elf-mcp-policy-lint checks the public package boundary: no private validation paths, no unrelated commercial-tool references, and no bundled solver outputs inside the public sample decks.

Bundled data (all generated from fresh ELF600 install via scripts/crawl_*.py):

  • help_dump.json — Shift_JIS HTM decoded + HTML-stripped
  • examples_dump.json — 228 MAGIC + 66 ELFIN + 38 BEAM input files
  • wiki_dump.json — 146 curated pages from https://elf.co.jp/
  • python_dump.jsonelftypes.py/magtypes.py (83 ctypes API functions each), *.cfg, ELFERR.def/MESERR.def, etc.

Curated topics (elf_usage)

Returns documentation on:

  • File formats: .mai (analysis input), .mei (mesh script), .meg (compiled mesh)
  • Solvers: MAGIC (magnetostatic, transient, AC), ELFIN (electrostatic), BEAM (particle tracking)
  • Eddy current: MAB / MAT / MBB elements, time-stepping, sinusoidal AC (SOL MOMC)
  • Element types: full catalog with DOF counts and symmetry restrictions (3D / 2D / Axisym)
  • B-H curves: anisotropy (HBA1/HBA2), recoil, extrapolation
  • Motor workflows: IPM Ld/Lq plus a radia-mcp/open-FEA concept bridge for air-gap field, torque, flux linkage/back-EMF, lamination, and eddy-current studies
  • Inductance: Lsc (JIS) and Ll (IEEJ) with 6 samples
  • Magnetization / demagnetization (MAGNE2)
  • Convergence troubleshooting, error codes (160+ ELF-Q/E/W codes)
  • Force methods: FORC vs FORT vs FIXB
  • Tools: IEmesh, Wmap3, MagFilter2, MaiEditor3, ELF/Bench

Available topics:

all, overview, mai_format, mei_format, meg_format,
magic, elfin, beam, element_types, bh_curves,
sol_commands, mei_commands, ipm_motor, motor_radia_bridge, inductance,
magnetization, examples, meg_export, treasure_box,
sinusoidal, anisotropy, sted, meshing, convergence,
force_methods, errors, iemesh, tools, cln_extraction,
licensing, python_api, live_drive

The cln_extraction topic documents the 6-step ELF MAGIC -> Cauer Ladder Network synthesis workflow (Foster fit + Cauer-I/II + 3-way validation against step response / Joule loss / Lorentz force). Distilled from a 21-script rectangular-CLN reference analysis suite.


Installation

pip install ELF-mcp-server

The current PyPI distribution name is ELF-mcp-server. The installed console scripts remain lowercase: elf-mcp-server and elf-mcp-policy-lint.

Verify:

elf-mcp-server --selftest

Usage

Claude Code

claude mcp add elf "C:/Program Files/Python312/Scripts/elf-mcp-server.exe"

(Adjust path for your Python install. On Linux/macOS, the script is typically ~/.local/bin/elf-mcp-server or similar.)

Cursor / Other MCP clients

Add to your MCP config:

{
  "mcpServers": {
    "elf": {
      "command": "elf-mcp-server"
    }
  }
}

Self-test

elf-mcp-server --selftest

Iterates through all curated topics and asserts non-empty documentation.


What is ELF600?

ELF600 is a commercial BEM (Boundary Element Method) electromagnetic analysis suite distributed by Science Solutions International Laboratory (https://www.science-solutions.jp/elf/).

Module Purpose
MAGIC Magnetostatic field (static, transient, AC). Eddy current via MAB/MAT/MBB.
ELFIN Electrostatic field analysis (D-E curves)
BEAM Charged particle beam tracking

Workflow:

.mei (mesh script) --> IEmesh --> .meg (compiled mesh)
.mai (analysis) + .meg --> MAGIC/ELFIN/BEAM --> .mag/.mao results

Why this server?

LLM coding agents authoring ELF input files (.mai/.mei) need access to:

  • The ~60k character ELF reference manual content,
  • Element naming conventions (T/K/R symmetry × element family),
  • SOL block recipes (MOME / MOMC / FIEL / FORC / NONL),
  • Frequency-sweep AC analysis structure,
  • Common error code interpretation,

without polluting context with the entire vendor PDF. This MCP server returns just the relevant topical chunk on demand.


License

BSD-3-Clause. See LICENSE.

ELF600 itself is a commercial product of Science Solutions International Laboratory and is not redistributed by this package — only documentation references.


Author

Kengo Sugahara, Kindai University (ksugahar@ele.kindai.ac.jp)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

elf_mcp_server-1.41.0.tar.gz (2.7 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

elf_mcp_server-1.41.0-py3-none-any.whl (3.0 MB view details)

Uploaded Python 3

File details

Details for the file elf_mcp_server-1.41.0.tar.gz.

File metadata

  • Download URL: elf_mcp_server-1.41.0.tar.gz
  • Upload date:
  • Size: 2.7 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for elf_mcp_server-1.41.0.tar.gz
Algorithm Hash digest
SHA256 b0e25cbff7766a74d78d85d7005758c753979308a59f411f7f4a589d002e5706
MD5 16bee69a3d36a23b525fd816a3c6b84f
BLAKE2b-256 b1f99e606ccd11f022f47af987e7df2f9e125c063cbc0b474726992533d54065

See more details on using hashes here.

Provenance

The following attestation bundles were made for elf_mcp_server-1.41.0.tar.gz:

Publisher: publish.yml on ksugahar/ELF-mcp-server

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file elf_mcp_server-1.41.0-py3-none-any.whl.

File metadata

File hashes

Hashes for elf_mcp_server-1.41.0-py3-none-any.whl
Algorithm Hash digest
SHA256 fa933d6dc19434c56797fa6d242ce946404dfdce91741e384d4736d512dd1b6f
MD5 43dca9a54139e741ce5f5f9d03369a2b
BLAKE2b-256 00b74def9c3c9976be699402c2ea4270bd64fcf05590aa8528f46cf779003f39

See more details on using hashes here.

Provenance

The following attestation bundles were made for elf_mcp_server-1.41.0-py3-none-any.whl:

Publisher: publish.yml on ksugahar/ELF-mcp-server

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page