spryn

Spryn is an open-source software tool for building and executing reproducible computational workflows. It provides a declarative workflow language that describes tasks, inputs, outputs, and dependencies, compiling these into a directed acyclic graph (DAG) of steps. The runtime can execute tasks on local machines, compute clusters, or cloud platforms using pluggable backends, enabling cross-environment portability.

Key features include a content-addressable cache that skips re-computation when inputs have not changed, automatic dependency resolution, and built-in support for containerization via Docker and Singularity. Spryn aims to be language-agnostic, with adapters for common languages and shells, allowing users to orchestrate data processing, simulation, and analysis pipelines. The configuration is designed to be human-readable and version-controllable.

Spryn emerged from a community-driven initiative in the late 2010s, with the first public release around 2020. Development is coordinated by a foundation/consortium and relies on open-source licensing, public issue trackers, and collaborative reviews. Over successive releases, Spryn added modular backends, improved caching, and richer logging and provenance features to enhance reproducibility.

Spryn is used in scientific research, data analysis, and data engineering to manage complex pipelines. Users value its simplicity relative to larger workflow systems and its emphasis on reproducibility, though it faces competition from established tools and may require more setup for very large graphs.

Related systems include Snakemake, Nextflow, and Airflow.