foam-agent

Create and inspect lattice Boltzmann flow simulations in the Foam-Agent Harness workspace.

465|41|Updated Aug 4, 2026
One-click install
npx skills add https://github.com/autonomous-ai/openharness --skill foam-agent-autonomous-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: foam-agent
Source: https://github.com/autonomous-ai/openharness/tree/main/store/agents/foam-agent/skills/foam-agent
Command: npx skills add https://github.com/autonomous-ai/openharness --skill foam-agent-autonomous-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Running and comparing fluid flow simulations normally requires deep CFD tooling expertise; this Skill lets you create, run, and inspect exploratory flow simulations inside the Foam-Agent Harness workspace with controlled parameters and reproducible results. ## Core Features & Use Cases - Simulation Runs: Execute a two-dimensional D2Q9 lattice Boltzmann solver via the studio toolchain, with results stored in out/runs/<id>/ and tracked through out/latest.json. - Flow Case Comparison: Save a baseline, vary shape or radius while holding inlet speed, viscosity, and solver steps fixed, then inspect flow.csv and field.json for velocities and convergence records. - Optional OpenFOAM Integration: Reproduce the supplied lid-driven cavity benchmark with blockMesh and icoFoam after reading the upstream foam.md instructions. - Use Case: An engineer exploring how inlet wind strength affects flow around a shape can adjust studio.json controls, run the solver, and compare the measured artifacts across runs while distinguishing preview results from converged engineering analysis. ## Quick Start Ask the agent to run a new flow simulation with the current studio.json controls and report which run contains the result and what was measured.

Frequently Asked Questions about foam-agent

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

FAQPage Schema
How do I run a flow simulation in the Foam-Agent workspace?

Run the studio toolchain with "$STUDIO_TOOLCHAIN/run.sh" simulate after reviewing controls in studio.json. Successful artifacts and measurements appear in out/runs/<id>/, and out/latest.json names the current result.

How do I compare flow cases across simulation runs?

Save a baseline run, then vary shape or radius while holding inlet speed, viscosity, and solver steps fixed. Read flow.csv and field.json to check finite velocities, solid-boundary velocities, and the convergence record.

Does this skill support real OpenFOAM simulations?

Yes, an optional openfoam action reproduces the supplied lid-driven cavity benchmark using blockMesh and icoFoam. Read $STUDIO_UPSTREAM/.claude/skills/foam.md first and preserve the benchmark boundary conditions.

What are the limitations of the local lattice Boltzmann solver?

The local workflow is a coarse two-dimensional D2Q9 exploratory model, not a converged engineering drag analysis. A coarse flow picture does not establish drag values or mesh convergence.

What happens when a simulation run fails?

A failed run preserves the last successful result and records the error in the verdict. The previous success remains available in out/runs/ so no valid data is lost.