pegasus-help

name	pegasus-help
description	Show available Pegasus workflow skills and which one to use
allowed-tools	["Read"]

Pegasus Workflow Toolkit — Navigation

You are a Pegasus workflow development assistant. The user has invoked /pegasus-help.

Display the following navigation table so the user knows which skill to use:

Available Skills

Skill	When to Use	What It Does
/pegasus-scaffold	Starting a new workflow from scratch	Generates a complete project: workflow_generator.py, wrapper scripts, Apptainer definition file, README, and manual test script
/pegasus-wrapper	Adding a single pipeline step	Generates a Python or shell wrapper script for one tool
/pegasus-apptainer	Building the container image	Generates an Apptainer .def file for your workflow's tool stack
/pegasus-convert	Migrating from Snakemake or Nextflow	Converts an existing pipeline definition to Pegasus
/pegasus-debug	Workflow failed and you need help	Diagnoses failures from Pegasus error logs and proposes fixes
/pegasus-review	Workflow is written but untested	Reviews a workflow for common pitfalls and best practices

Reference Materials

These live in the pegasus-ai plugin directory — the one holding skills/, references/ and assets/ — which is not your working directory. If the context you were given does not name it, find it with a glob for **/pegasus-ai/references/PEGASUS.md.

• references/PEGASUS.md — Comprehensive guide covering all Pegasus concepts, patterns, and pitfalls
• assets/templates/ — Copy-paste-and-customize starting points for all file types
• assets/examples/ — Curated reference files from 12 production workflows

Example Workflows

These production workflows are included in assets/examples/ and available as full repositories:

Example	Key Patterns	Full Repository
workflow_generator_tnseq.py	Per-sample parallelism, fan-in merge, R/JAR support files	pegasus-isi/tnseq-workflow
workflow_generator_earthquake.py	API data fetch, per-region loops, no replica catalog inputs	pegasus-isi/earthquake-workflow
workflow_generator_mag.py	Shell wrappers, is_stageable=False, micromamba, --test mode, skip flags	pegasus-isi/mag-workflow
workflow_generator_soilmoisture.py	ML train-then-predict, per-polygon parallelism	pegasus-isi/soilmoisture-workflow
workflow_generator_airquality.py	Dual pipeline, skip flags, multiple data sources, fan-in merge	pegasus-isi/airquality-workflow
workflow_generator_gwas_qc.py	Fork-join topology, PLINK bioinformatics, complex branching	pegasus-isi/gwas-qc-workflow
workflow_generator_rnaseq.py	Nextflow conversion, R support files (edgeR/DESeq2), per-sample	pegasus-isi/rnaseq-workflow
workflow_generator_proteinfold.py	GPU protein folding, CondorIO for model caches, batch inference	pegasus-isi/proteinfold-workflow
workflow_generator_s2_segmentation.py	Image tiling, split→parallel→merge, GPU U-Net training	pegasus-isi/s2-segmentation-workflow
workflow_generator_medical_imaging_fl.py	Federated learning with SubWorkflows, FL rounds as sub-DAGs	pegasus-isi/medical-imaging-fl-workflow

Additional Workflow Repositories

These workflows are also available but share patterns with the examples above:

Workflow	Key Patterns	Full Repository
crophealth	CNN image classification, edge-to-cloud DPU	pegasus-isi/crophealth-workflow
orcasound	S3 data fetch, per-sensor parallelism, hydrophone audio	pegasus-isi/orcasound-workflow
seaice	NASA Earthdata API, LSTM/MLP classifier, satellite geospatial	(in development)
sprite	Federated learning, tar archive data transfer, per-site parallel	(in development)

Quick Start

If you're creating a new workflow, start with /pegasus-scaffold.

If you're modifying an existing workflow, use /pegasus-wrapper (to add a step), /pegasus-review (to check for issues), or /pegasus-debug (to fix failures).