Skip to content
Back to skills

Solver Benchmark

ASecurity

Run N-way diffraction solver benchmarks comparing AQWA, OrcaWave, and

  • 17 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 9, 2026
testingpythongobashapidocumentation

Works with

  • api

Security analysis

A100/100

Scanned September 9, 2026

npx -y skills add vamseeachanta/workspace-hub --skill solver-benchmark --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Solver Benchmark?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Solver Benchmark
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vamseeachanta-solver-benchmark/badge)](https://www.skillsdirectory.com/skills/vamseeachanta-solver-benchmark)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: solver-benchmark
version: 2.0.0
description: Run N-way diffraction solver benchmarks comparing AQWA, OrcaWave, and
  BEMRosetta results
type: reference
author: workspace-hub
category: engineering-utilities
tags:
- benchmark
- aqwa
- orcawave
- bemrosetta
- diffraction
- comparison
- validation
platforms:
- engineering
invocation: /benchmark-solvers
capabilities: []
requires: []
---

# Solver Benchmark

## When to Use This Skill

Use solver benchmarking when:
- **Validating solver setups** - Confirm all solvers produce consistent results
- **Cross-validating results** - Compare different solver outputs for the same hull
- **Identifying solver quirks** - Find solver-specific behaviors or numerical differences
- **Quality assurance** - Verify analysis methodology across tools
- **Reporting** - Generate comparison reports for documentation

## Quick Start

```bash
# Run 3-way benchmark on Unit Box (validation case)
uv run python scripts/benchmark/run_3way_benchmark.py \
    specs/modules/benchmark/unit_box_spec.yml

# Dry run (generate files, skip solver execution)
uv run python scripts/benchmark/run_3way_benchmark.py \
    specs/modules/benchmark/unit_box_spec.yml --dry-run

# Run specific solvers only
uv run python scripts/benchmark/run_3way_benchmark.py \
    spec.yml --solvers orcawave,aqwa

# Custom output directory
uv run python scripts/benchmark/run_3way_benchmark.py \
    spec.yml -o results/my_benchmark
```

## Related Skills

- **mesh-utilities** - Mesh inspection and conversion (`/mesh`)
- **hydrodynamic-analysis** - BEM theory and RAO analysis
- **orcaflex/specialist** - OrcaFlex integration

## Sub-Skills

- [Troubleshooting](troubleshooting/SKILL.md)

## Sub-Skills

- [Solver Paths](solver-paths/SKILL.md)
- [Benchmark Workflow](benchmark-workflow/SKILL.md)
- [Output Structure](output-structure/SKILL.md)
- [Consensus Levels](consensus-levels/SKILL.md)
- [Creating Benchmark Specs](creating-benchmark-specs/SKILL.md)
- [Available Hull Geometries](available-hull-geometries/SKILL.md)
- [Programmatic Usage](programmatic-usage/SKILL.md)
- [MultiSolverComparator (+1)](multisolvercomparator/SKILL.md)
- [OrcaWave (via OrcFxAPI) (+3)](orcawave-via-orcfxapi/SKILL.md)
- [Pre-Flight Validation Checklist](pre-flight-validation-checklist/SKILL.md)
- [Correlation Thresholds (+1)](correlation-thresholds/SKILL.md)
- [Mandatory Checks Before Committing Reports (+2)](mandatory-checks-before-committing-reports/SKILL.md)
- [Related Work Items](related-work-items/SKILL.md)

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.