DataLad Hub

Overview #

DataLad Hub is a public deployment of Forgejo-Aneksajo – the Forgejo fork with native git-annex support. It provides a web interface for publishing, browsing, cloning, and collaborating on DataLad datasets without the need to deploy and maintain your own infrastructure.

DataLad Hub is not a separate tool or codebase – it is an instance of Forgejo-Aneksajo, much like github.com is an instance of GitHub’s proprietary forge. The underlying technology is described on the Forgejo-Aneksajo page.

Key Features #

• Dataset hosting – push DataLad datasets including both git metadata and git-annex content
• Web browsing – explore dataset contents, file trees, and metadata through the Forgejo web interface
• Standard git workflows – clone, fork, pull request, and collaborate using familiar git patterns
• git-annex support – full git-annex protocol support means datalad push and datalad get work seamlessly
• Organizations and teams – group datasets by lab, project, or collaboration with appropriate access controls

git-annex / DataLad Integration #

Integration level: native-datalad.

# Create a sibling on DataLad Hub
datalad create-sibling-gogs --name hub \
    --api https://hub.datalad.org/api/v1 \
    --credential datalad-hub-token

# Push dataset (git refs + annexed content)
datalad push --to hub

Because the backend is Forgejo-Aneksajo, both git and git-annex content are handled by the same server. There is no need to configure separate special remotes for annexed content.

Others to Consider #

datalad-registry (live at registry.datalad.org) – a service for auto-discovery and metadata extraction of DataLad datasets. Rather than hosting datasets, it indexes datasets discovered on GitHub and other hosts, extracting metadata to make them searchable. Could be a useful complement to DataLad Hub for making archived datasets discoverable.

AI Readiness #

Level: ai-partial.

The API provides programmatic access to repository metadata, file listings, and issue discussions – all structured and AI-consumable. The actual dataset content depends on the specific datasets hosted and may require domain-specific processing.

Relationship to Lab-in-a-Box #

DataLad Hub can be thought of as a managed Forgejo-Aneksajo deployment focused on dataset hosting. Research groups that want the same capabilities on their own hardware can deploy Lab-in-a-Box instead.

See Also #

• Forgejo-Aneksajo – the technology powering DataLad Hub
• Lab-in-a-Box – self-hosted alternative