The community schema manages organisations and users. Community management in the AI4SH database relies on the default framework processes that are registered by the process/processes_records_v10_sql.json file during the database setup stage. No separate community process registration files are included in the AI4SH setup_processes workflow.

What the framework provides by default

After running processes_records_v10_sql.json, the framework has built-in processes for:

  • Adding organisations (community.organisation)
  • Adding user categories (community.user_categories)
  • Adding users (community.user)
  • Updating user details
  • Managing user activity

These bootstrap processes are part of the core framework and do not need to be registered again for AI4SH.

Managing community data after setup

Once the database is running, community records (organisations and users) are managed through scheme files and process files referencing the default community processes. To add a new user:

  1. Create a process file calling manage_user (or the equivalent community process)
  2. Supply the required parameters: organisation_id, user_name, password, stratum_code, and any optional personal details
  3. Run the process via the appropriate notebook

The stratum_code determines which .env credentials the user logs in with and thus which database privileges they have. The 8 default strata range from 0 (most restricted) to 5 (most permissive), plus community_admin and login_evaluation for administrative functions.

Cross-schema relationship

Community records are referenced by:

  • observation.data_source — simplified passive data supplier (not a full framework user)
  • observation.person — non-registered individual attributed to data (GDPR consideration)

These two tables in the observation schema are intentionally decoupled from the full community.user table, allowing attribution of data to people who are not system users.

Updated: