The Common Workflow Language (CWL) is an informal, multi-vendor working group consisting of various organizations and individuals that have an interest in portability of data analysis workflows. Our goal is to create specifications that enable data scientists to describe analysis tools and workflows that are powerful, easy to use, portable, and support reproducibility.
CWL builds on technologies such as JSON-LD and Avro for data modeling and Docker for portable runtime environments.
CWL is designed to express workflows for data-intensive science, such as Bioinformatics, Chemistry, Physics, and Astronomy.
The current stable specification is draft 2:
http://common-workflow-language.github.io/draft-2/
Implementations capable of running the Common Workflow Language:
Github repository of example workflows.
The CWL effort is on Github:
https://github.com/common-workflow-language/common-workflow-language
There is a Jenkins server maintained by Curoverse that runs tests for the reference implementation, builds and uploads packages, and builds and uploads Docker images:
https://ci.curoverse.com/job/common-workflow-language/
Current build status:
If you are interested in contributing ideas or code, please join the mailing list or fork the repository and send a pull request!
- Curoverse
- Seven Bridges Genomics
- Galaxy
- Institut Pasteur
- BioDatomics
- Michigan State University
- Broad Institute
- Peter Amstutz peter.amstutz@curoverse.com
- John Chilton jmchilton@gmail.com
- Michael R. Crusoe mcrusoe@msu.edu
- John Kern kern3020@gmail.com
- Hervé Ménager herve.menager@gmail.com
- Maxim Mikheev mikhmv@biodatomics.com
- Tim Pierce twp@unchi.org
- Stian Soiland-Reyes soiland-reyes@cs.manchester.ac.uk
- Luka Stojanovic luka.stojanovic@sbgenomics.com
- Nebojsa Tijanic nebojsa.tijanic@sbgenomics.com