license: cc-by-4.0
configs:
- config_name: lock_file_snapshot
data_files:
- split: data
path: data/lock_file_snapshot.parquet
- config_name: repository
data_files:
- split: data
path: data/repository.parquet
- config_name: source_markdown_file_history
data_files:
- split: data
path: data/source_markdown_file_history.parquet
- config_name: source_markdown_file_snapshot
data_files:
- split: data
path: data/source_markdown_file_snapshot.parquet
- config_name: source_markdown_file_version
data_files:
- split: data
path: data/source_markdown_file_version.parquet
A Dataset of GitHub Agentic Workflow Histories: Early Adopters
Overview
GAWD is a dataset capturing the emergence of GitHub Agentic Workflows (GH-AW) among early-adopter open-source repositories. This release spans 262 repositories, 2820 observed GH-AW source markdown file snapshots, 2820 source-version records across 604 file histories, and 2820 associated lock-file snapshots.
Call for Papers
Researchers and practitioners are invited to leverage the GAWD dataset to advance research in Natural Language-Based Software Engineering. The dataset enables studies on how natural-language specifications, instructions, and workflow descriptions are used to define, execute, and evolve GitHub Agentic Workflows in real-world open-source repositories.
GAWD can support reproducible empirical studies, tool demonstrations, replication studies, negative results, and position papers aligned with NLBSE tracks on NLP methods, datasets, tools, and emerging challenges for software engineering.
NLBSE 2027 Workshop (https://nlbse2027.github.io)
Deadline: pending Location: Dublin, Ireland
Quick Look
The overview of the GAWD dataset is as follows:
| Dataset | #Repo | #Source Histories | #Source Versions | #Lock Snapshots |
|---|---|---|---|---|
| GAWD | 262 | 604 | 2,820 | 2,820 |
Intended Uses
Pending.
Dataset Schema
A detailed explanation about the fields of the tables can be found in Tables dictionary.
License
This dataset aggregates content from GitHub repositories. Each source repository retains its original copyright and license (e.g., MIT, Apache-2.0, GPL family, Creative Commons variants, etc.). Files, histories, and any other artifacts originating from those repositories remain governed by their original licenses.
Users must verify and comply with the specific license of any source repository or file they extract or use from this collection. Do not assume a universal re-license. If you believe content appears here in a way that conflicts with its license, please contact the maintainers, and it will be removed.
Citation
Citation metadata is pending. Please cite the release version and dataset URL.