| name | create-hydraulic-model |
| description | Create Modelica models using the Hydraulic library with knowledge-graph-guided component selection and wiring. Use this skill whenever the user asks to create, build, design, or generate a hydraulic circuit or Modelica model using the Hydraulic library. Triggers on phrases like 'create a hydraulic model', 'build a circuit with a pump and cylinder', 'design a meter-in circuit', 'make a model with a directional valve', 'generate a hydraulic system', or any request to assemble Hydraulic library components into a working model. For the overall structure of a large model or library, consult modelica-model-architecture first. |
Create Hydraulic Model (Graph RAG Guided)
This skill uses a knowledge graph of the Hydraulic Modelica library to guide the user through creating valid hydraulic circuit models. The graph (built from Hydraulic 2.1) contains 168 parsed components with their ports, parameters, inheritance, and 60 validated connection patterns extracted from the library's composite models and 18 example circuits.
Before you run anything
The query commands below run python -m Hydraulic.main …; the optional validation step
(Step 7) additionally drives the shared launcher. For the Windows-vs-Unix shell/Python
rules (PowerShell on Windows, python vs python3) and launcher conventions, see
the shared-conventions appendix at the end of this file — the launcher-specific parts (temp dirs,
JSON output, MSL, --load-library) only matter once you validate.
Knowledge Graph (bundled)
The Graph RAG dataset and query module ship inside this skill as the
Hydraulic/ Python package — there is nothing external to locate. The query
path reads the prebuilt graph in Hydraulic/data/graph.json; it does not
need the Modelica library and does not re-parse anything, so it works wherever
the skill is installed.
- Requirement:
networkx — but you don't install it by hand. Hydraulic.main
self-provisions it into a managed venv on first run (it never touches system
Python). To pre-warm: python3 "<scripts-dir>/bootstrap_env.py" networkx.
- Run the query module with this skill's own directory (the folder
containing this
SKILL.md) as the working directory, so the package resolves
as Hydraulic.
Query Commands
Run from this skill's directory (<skill-dir> below — the folder that contains
this SKILL.md) so the Hydraulic package is importable. Use python instead
of python3 on Windows if that's what's on PATH.
cd "<skill-dir>"
python3 -m Hydraulic.main --query "pump"
python3 -m Hydraulic.main --details "CylinderDouble"
python3 -m Hydraulic.main --interface "FourPort"
python3 -m Hydraulic.main --examples
python3 -m Hydraulic.main --rag "pressure control circuit"
(Use the absolute path to this skill's directory for <skill-dir>.)
Validating & Simulating the Generated Model
Generation itself only needs the bundled graph (above). To flatten, validate, or
simulate a generated model you drive WSMKernelX through the shared launcher — the same
one every other Modelica skill uses. Read
the shared-conventions appendix at the end of this file first for launcher resolution, the
Windows-vs-Unix shell/Python rules, the temp-dir/cleanup conventions, the JSON-array
output gotcha, and the MSL 4.x dialect notes.
The model extends Hydraulic.…, so the Hydraulic library must be loaded alongside it.
Do not locate or pass the path by hand — add --load-library Hydraulic to any
validate / simulate / diagnose run and the launcher finds it (pin a version with
Hydraulic==2.1; run --mode libraries to see what is installed). Resolution
details and overrides:
Appendix → Using non-MSL libraries.
Mind the library version. The bundled graph is built from Hydraulic 2.1. If
--mode libraries shows a different installed version, treat graph answers as hints
rather than ground truth — component or parameter names may have drifted — and validate
the generated model early so any mismatch surfaces immediately.
Guided Workflow
Follow these steps in order, confirming with the user at each step before proceeding.
Step 1: Understand the Circuit Requirements
Parse the user's request (from $ARGUMENTS if invoked as /create-hydraulic-model <desc>, or from the conversation).
Identify:
- Primary function: What should the circuit do? (move a load, control flow, hold pressure, transmit power, etc.)
- Actuator type: Translation (cylinder) or rotation (motor)?
- Control needs: Directional control? Pressure limiting? Flow control? Sequence?
- Any specific components the user mentioned by name
Step 2: Query the Knowledge Graph for Relevant Components
Based on the requirements, run targeted queries:
python -m Hydraulic.main --details "CylinderDouble"
python -m Hydraulic.main --details "Motor"
python -m Hydraulic.main --details "Pump"
python -m Hydraulic.main --interface "FourPort"
python -m Hydraulic.main --interface "ThreePort"
python -m Hydraulic.main --interface "TwoPortStatic_2"
python -m Hydraulic.main --rag "<user's circuit description>"
Also read the entities.json directly to get parameter defaults:
python -c "
import json
with open('Hydraulic/data/entities.json') as f:
entities = json.load(f)
for e in entities:
if e['name'] == 'COMPONENT_NAME':
print(json.dumps(e, indent=2))
break
"
Step 3: Present Component Selection to User
Show the user a table of recommended components:
Proposed Components:
| Role | Component | Description |
|-------------------|------------------------------------|--------------------------------|
| Power source | Hydraulic.PumpsAndMotors.Pump | Fixed displacement pump |
| Actuator | Hydraulic.Cylinders.CylinderDouble | Double-acting cylinder |
| Direction control | Hydraulic....PCVE43ClosedCenter | 4/3 proportional valve |
| Pressure relief | Hydraulic....PressureReliefValve | System pressure limiter |
| Tank (supply) | Hydraulic.LiquidContainers.Tank | Constant pressure reservoir |
| Tank (return) | Hydraulic.LiquidContainers.Tank | Return line reservoir |
| Speed source | Modelica...ConstantSpeed | Drives the pump shaft |
| Load | Modelica...Mass | Translational mass load |
| Control signal | Modelica.Blocks.Sources.Step | Step input for valve command |
Ask the user:
- "Does this component selection look right?"
- "Would you like to add, remove, or swap any components?"
- "Any specific valve type preference?" (show available options if relevant)
Wait for user confirmation before proceeding.
Step 4: Propose Connections
Query example circuits that use similar components to find validated connection patterns:
python -m Hydraulic.main --rag "connect pump cylinder valve"
Present the proposed wiring to the user. Group by domain:
Proposed Connections:
Hydraulic (blue):
1. tank1.port -> pump.port_a (tank feeds pump inlet)
2. pump.port_b -> controlValve.port_p (pump pressure to valve P port)
3. controlValve.port_a -> cylinder.port_a (valve A to cylinder A)
4. controlValve.port_b -> cylinder.port_b (valve B to cylinder B)
5. controlValve.port_t -> tank2.port (valve T to return tank)
6. reliefValve.port_a -> pump.port_b (relief across pump output)
7. reliefValve.port_b -> tank3.port (relief to tank)
Mechanical (green):
8. constantSpeed.flange -> pump.flange (speed source drives pump)
9. cylinder.flange_b -> mass.flange_a (cylinder moves mass)
Signal (blue):
10. step.y -> controlValve.u1 (command signal to valve)
11. const.y -> controlValve.u2 (zero signal to valve u2)
Ask the user: "Does this wiring look correct? Any connections to add or change?"
Wait for user confirmation before proceeding.
Step 5: Configure Parameters
For each component, show key parameters with defaults. Read these from the knowledge graph:
python -m Hydraulic.main --details "Pump"
python -m Hydraulic.main --details "CylinderDouble"
Present a parameter table:
Key Parameters (defaults shown — override any you want):
Pump:
D = 5e-05 [m^3/rev] "Displacement"
CylinderDouble:
diameterPiston = 0.05 [m]
diameterRod_a = 0.005 [m]
lengthHousing = 0.5 [m]
lengthPiston = 0.05 [m]
lengthRod_a = 0.5 [m]
Vdead_a = 1e-05 [m^3]
mPiston = 10 [kg]
PressureReliefValve:
pMin = 20000000 [Pa] "Opening pressure (200 bar)"
Mass:
m = 100 [kg]
ConstantSpeed:
w_fixed = 100 [rad/s]
Simulation:
StopTime = 2 [s]
Ask: "Want to change any parameter values?"
Wait for user confirmation before proceeding.
Units. Hydraulic models are SI (pressure in Pa, flow in m³/s) and use displayUnit for
friendly labels (e.g. bar, L/min):
- A parameter modifier that carries
displayUnit must take a literal value:
pMin(displayUnit = "bar") = 20000000.0, not = 200*1e5. An arithmetic expression silently
disables the displayUnit conversion in WSM, so the GUI shows the raw SI number. Compute the SI
value yourself and write it as a literal (keep the readable origin in a comment if helpful).
- Signals arriving through a connector are already in SI base units — never rescale at the wiring
site (
*1e-5, /60000). Declare SI types with displayUnit = "…" and let the unit conversion
happen for display only.
Step 6: Generate the .mo File
Ask the user where to save the model:
- Suggested default: an
Examples/ subfolder of the user's own Modelica library
- Or any user-specified path
Generate the complete Modelica model following this template structure:
within <package.path>;
model <ModelName> "<Description>"
extends Hydraulic.Icons.Example;
extends Hydraulic.Media.BaseModel;
// Components — pass medium = medium only if <Type> extends Hydraulic.Media.BaseModel (e.g. not Tank)
<Type> <instanceName>(<params>, medium = medium)
annotation(Placement(visible = true, transformation(origin = {x, y},
extent = {{-10, -10}, {10, 10}}, rotation = 0)));
// Parameters
parameter SI.Pressure openingPressure = 20000000.0 "Relief valve opening pressure";
equation
// Hydraulic connections
connect(<inst1>.<port>, <inst2>.<port>)
annotation(Line(visible = true, origin = {0, 0},
points = {{0, 0}, {0, 0}}, color = {0, 170, 255}));
// Mechanical connections
connect(<inst1>.<flange>, <inst2>.<flange>)
annotation(Line(visible = true, origin = {0, 0},
points = {{0, 0}, {0, 0}}, color = {0, 127, 0}));
// Signal connections
connect(<source>.y, <target>.u1)
annotation(Line(visible = true, origin = {0, 0},
points = {{0, 0}, {0, 0}}, color = {0, 0, 127}));
annotation(
experiment(StopTime = <stopTime>, __Wolfram_NumberOfIntervals = 2000,
__Wolfram_Algorithm = "cvodes"),
Documentation(info = "<html><body><Description></body></html>"));
end <ModelName>;
Critical rules for valid .mo files:
- Every Hydraulic component that extends
Hydraulic.Media.BaseModel needs medium = medium in its parameter list; components that don't (e.g. Tank) must NOT receive it — check with --details <Name>
extends Hydraulic.Media.BaseModel; provides the medium record
extends Hydraulic.Icons.Example; gives the example icon
- Use color
{0, 170, 255} for hydraulic connections
- Use color
{0, 127, 0} for mechanical connections
- Use color
{0, 0, 127} for signal connections
- Every
connect() needs a Line annotation (even with dummy points)
- The
within path must match the file's location in the package hierarchy
- Pump needs
phi.fixed = true, dp.fixed = true for proper initialization
- CylinderDouble needs
pA.fixed = true, pB.fixed = true for initialization
Layout conventions (approximate placement origins for a clean diagram):
- Tanks: bottom row, y = -70
- Pump: left side, y = -40
- Relief valve: near pump, y = -40
- Directional valve: center, y = 10
- Cylinder: upper area, y = 50
- Mass/load: right side, y = 50
- Signal sources: far left, various y
- Speed source: far left, y = -40
Write the file using the Write tool.
Step 7: Offer Validation
After writing the file, ask: "Would you like me to validate this model using the Modelica compiler?"
If yes, use the validate-modelica (or simulate-modelica) skill on the generated file.
Because the model uses the Hydraulic library, add --load-library Hydraulic so the
launcher loads it alongside the model (MSL is automatic):
python3 "<scripts-dir>/wsm_run.py" --mode validate \
--model "<path-to-ModelFile.mo>" --name <ModelName> --load-library Hydraulic --timeout 120
(Use python on Windows. See "Validating & Simulating the Generated Model" above and
the shared-conventions appendix for launcher resolution and temp-dir/cleanup
conventions.) A clean run reports flatten=Pass.
Component Quick Reference
Common Circuit Patterns
Basic servo (translate): Pump + PressureReliefValve + PCVE43* + CylinderDouble + Mass
Basic servo (rotate): Pump + PressureReliefValve + PCVE43* + Motor + Inertia
Meter-in flow control: Add FixedTurbulentThrottle before directional valve
Meter-out flow control: Add FixedTurbulentThrottle after directional valve
Load holding: Add CounterBalanceValve between valve and cylinder
Accumulator circuit: Add CheckValve + OneWayFlowControlValve + GasChargedAccumulator
Sequence circuit: Add PilotOperatedSequenceValve between stages
Pressure reducing: Add PressureReducingValve for sub-circuit pressure limiting
Valve Naming Convention
- DCVE: Directional Control Valve, Electrically actuated (conventional solenoid)
- PCVE: Proportional Control Valve, Electrically actuated (proportional solenoid)
- DCVH: Directional Control Valve, Hydraulically actuated (pilot operated)
- DCVM: Directional Control Valve, Mechanically actuated
- 42: 4 ports, 2 positions
- 43: 4 ports, 3 positions
- 32: 3 ports, 2 positions
- 63: 6 ports, 3 positions
- Center type suffix: ClosedCenter, OpenCenter, TandemCenter, FloatingCenter, DiagonalCenter
Port Naming Convention
- port_p: Pressure supply (from pump)
- port_t: Tank return
- port_a: Load port A (cylinder side A)
- port_b: Load port B (cylinder side B)
- port_pilot: Pilot pressure input
- u1, u2: Electrical/signal control inputs
- flange, flange_a, flange_b: Mechanical connections
- port: Single hydraulic port (tanks, accumulators, sensors)
Required Modelica Standard Library Components
Modelica.Mechanics.Rotational.Sources.ConstantSpeed — drives pump shaft
Modelica.Mechanics.Translational.Components.Mass — translational load
Modelica.Mechanics.Rotational.Components.Inertia — rotational load
Modelica.Blocks.Sources.Step — step signal for valve command
Modelica.Blocks.Sources.Constant — constant signal (e.g., k=0 for valve u2)
Modelica.Blocks.Sources.Ramp — ramp signal
Modelica.Blocks.Math.Min / Max / Add — signal processing
Modelica.Blocks.Math.BooleanToReal — convert switch to proportional signal
Appendix: shared conventions for the Modelica skills
Shared by every Modelica skill that drives WSMKernelX through the
bundled launcher; inlined here at release time. For the CLI/option
reference, environment variables (WSM_HOME, WSM_VSDEVCMD),
install discovery, and the analysis scripts, see
../scripts/README.md.
Locating the launcher
<scripts-dir> (used throughout the skills) is the shared scripts/ folder.
Some installs symlink the skill directories without it, so resolve it in this
order and use the first that exists:
$WSM_SKILLS_SCRIPTS (bash) or $env:WSM_SKILLS_SCRIPTS (PowerShell), if set.
../scripts relative to the skill directory — in a normal install
../scripts/wsm_run.py already exists, so use that path directly; do not run
a shell probe to "resolve" it.
- The repo checkout you installed from, e.g.
.../agentskills/scripts.
- Last resort, search the home directory:
- PowerShell:
Get-ChildItem $HOME -Recurse -Filter wsm_run.py -ErrorAction SilentlyContinue | Select-Object -First 1
- bash/zsh:
find ~ -name wsm_run.py -path '*scripts*' 2>/dev/null | head -1
If only #4 finds it, the install is missing the scripts/ link — tell the user
to run install.sh (or install.ps1) from the repo, which links scripts/ too.
Shell and Python
On Windows, use PowerShell. The Git-Bash/cygwin layer may be broken (even
ls/find can be absent, giving a misleading "exit 127 / command not found").
Run wsm_run.py with python (not python3); those calls are single-line and
shell-agnostic. For cleanup use Remove-Item -Recurse -Force, not rm -rf.
On macOS/Linux any POSIX shell is fine and python3 is the usual name.
Let the launcher own .mos/.bat and paths
Do not hand-write .mos scripts, .bat files, or hardcode install/compiler
paths. The bundled scripts/wsm_run.py handles every OS difference — it finds
the System Modeler install and kernel binary (macOS / Windows / Linux), finds and
loads the right MSL files, generates the .mos, and runs the kernel with a
working compiler environment per platform (system clang/gcc on macOS/Linux; the
Visual Studio dev environment via VsDevCmd.bat on Windows).
See ../scripts/README.md for WSM_HOME, the Windows compiler prerequisites,
and the full option table.
When the install or compiler isn't found
The launcher searches each OS's standard install locations. If it prints
ERROR: Could not locate a Wolfram System Modeler installation, the install
is in a non-standard place — ask the user for it and re-run with
--wsm-home "<path>" (or have them set WSM_HOME).
Building and simulating also need a C++ toolchain:
- Windows: Visual Studio Build Tools. The launcher locates
VsDevCmd.bat
itself; if it reports the compiler environment is missing, pass
--vsdevcmd "<path-to-VsDevCmd.bat>" (or set WSM_VSDEVCMD) and make sure
Build Tools are installed.
- macOS: the Xcode command-line tools (
xcode-select --install).
- Linux: gcc/g++.
Run python3 "<scripts-dir>/wsm_run.py" --mode info to see what the launcher
discovered.
Temporary directories
The launcher works in a _wsm_<mode>_temp/ directory next to the .mo file
(_wsm_validate_temp/, _wsm_simulate_temp/, _wsm_diagnose_temp/) and leaves
its outputs there. Tell the user, e.g.: "Working in temporary directory
_wsm_<mode>_temp/. This will be deleted afterwards." Pass --tempdir
to reuse one directory across models in a session.
Clean up by removing the whole directory — use the user's shell:
rm -rf "<model-dir>/_wsm_<mode>_temp"
Picking the model name
- The user may provide a path to a
.mo file, or you may already be working with
one in context.
- Extract the model name: the identifier after
model on the first non-comment
line, e.g. model FooBar → FooBar. The filename does not always match the
model name — parse the actual model/package declaration.
- For packages or nested models, use the top-level model name.
- Pick an instantiable model, not a package, for any kernel call. A
package
cannot be validated or simulated ("Invalid instantiation … is a package") — use
a nested model's full dotted name, e.g. Package.Model.
- Pass an absolute path to
--model (relative paths break as the working
directory shifts between calls).
Directory-form (multi-file) libraries
A directory-form library stores one class per file with a package.mo at each
level. You cannot validate such a class by handing the launcher only its own
.mo file — the class's within Lib; clause needs the whole package loaded, and
loading the single file alone fails with
Internal error: ... expandLibNode: Unknown library: Lib. Instead point
--model at the library folder (or its top package.mo, or any class file
inside it) and pass the full dotted class name via --name:
python3 "<scripts-dir>/wsm_run.py" --mode validate \
--model "/abs/path/InvertedPendulum" \
--name InvertedPendulum.Controller
The launcher resolves any of those forms up to the library's root package.mo
and loads the entire package (following package.order) before instantiating
--name. It prints a NOTE: telling you which package.mo it loaded. Do not
try to work around the unknown-library error by --load-ing individual files.
Reading the JSON output
The kernel writes <mode>.out.json into the temp dir. It is a JSON array —
take the first element, then read:
status.flatten: "Pass" / "Fail" (the primary result for validate).
status.build: "Pass" / "Fail" — C++ compilation/linking (simulate).
status.result: simulation result status (simulate).
messages.errors / messages.warnings / messages.notifications:
arrays (empty if none).
flat_model (validate) / simulation.resultFile (simulate): the
flattened class / path to the .mat.
See ../scripts/README.md (wsm_run.py section) for the full field reference.
MSL 4.x dialect
This toolchain ships MSL 4.x. When authoring models, use the 4.x names — the
3.2 names flatten with confusing "not found" errors:
- units:
Modelica.Units.SI.* (not Modelica.SIunits.*)
- source frequency parameter:
f (not freqHz), e.g. SineVoltage(V=.., f=..)
- declare the dependency as
annotation(uses(Modelica(version="4.0.0")))
Run python3 "<scripts-dir>/wsm_run.py" --mode info to confirm the exact MSL
version. wsm_run.py also warns on stderr if it spots a 3.2 name in the model.
When a flatten fails with Element not found ... in Modelica..., the MSL
component path is wrong (a misremembered name, not a missing install). Resolve
the correct path with the search-modelica-docs skill — e.g. sine is
Modelica.Blocks.Sources.Sine (not Math.Sine), difference is Math.Feedback
(not Math.Subtract), and saturation is Nonlinear.Limiter (not
Nonlinear.Saturation). Do not grep or walk the System Modeler install tree
to hunt for the class.
Using non-MSL libraries (Hydraulic, and other installed libraries)
MSL is automatic (--msl). For any other library a model uses — Hydraulic, or
anything the user installed from the Library Store — the launcher can find it for you.
Two steps, no guessing at paths:
-
See what is installed (bundled with System Modeler, user-installed archives, and
any custom folders configured in Model Center):
python3 "<scripts-dir>/wsm_run.py" --mode libraries
To get just one library's package path (e.g. for a manual --load):
--mode libraries --library Hydraulic (add --library-version 2.1 to pin a version).
-
Load it into a build by name — add --load-library <Name> to a validate /
simulate / diagnose run (repeatable; pin a version with Name==Ver):
python3 "<scripts-dir>/wsm_run.py" --mode validate \
--model "<path>/M.mo" --name M --load-library Hydraulic
The launcher resolves the library (bundled → user-installed → Model-Center custom
path, newest version wins) and loadFiles it before the model. A library usually
pulls in MSL, so keep MSL on (auto, or --msl yes). Override a lookup with
$WSM_LIBRARY_<NAME> (e.g. WSM_LIBRARY_HYDRAULIC=/path/to/package.moe), or fall
back to an explicit --load <path-to-package.mo|.moe>.