Database Access lets you connect modern business intelligence (BI) tools—such as Metabase, Tableau, or Power BI—or any SQL client directly to your organization's DevStride data. You generate a set of read-only PostgreSQL credentials, plug them into your tool of choice, and build custom dashboards, cross-reference metrics, and generate insights beyond the reports built into DevStride.
When you generate credentials, DevStride creates a dedicated read-only PostgreSQL role scoped to your organization. The connection points at a read replica of the database (not the live primary), so your queries never affect production performance. Access is limited to SELECT only, and row-level security policies keep every query isolated to your own organization's data.
The page header reads Manage Database Access.
The rail only shows what your role can manage, so if the entry is missing you don't have Manage database credentials. Ask an Owner or Admin to grant it to your role.
If no credentials exist yet, the page shows a Database Access card explaining that you can generate read-only database credentials to access your organization's data and connect using any SQL client or BI tool.
The Connection Details modal gives you two ways to copy your connection information:
postgresql)5432)public)The modal also suggests a display name for your BI tool, formatted as DevStride - {Your Organization Name}, with a Copy button so the connection is easy to recognize in your tool.
Once a credential is active, the card shows the role name, a colored Status label, the date it was created, and a row of action buttons. The active section is titled Read-Only Database Access.
As credentials are provisioned and managed, you may see any of these statuses:
SELECT access only—they can't change your data.DevStride is renaming its status and workstream vocabulary throughout the product — the API and MCP tools already use the new words, with deprecation windows listed on the API keys page — and the read-only database will follow in a later release: the tables lanes, lane_collections and folders become statuses, status_collections and workstreams, and the kept tables gain canonical twins of their lane_* and *_folder_* columns. Nothing changes for your queries on the day of that release: the current table names stay available as read-only views with identical columns, and the current column names keep working alongside their twins.
The legacy names are removed no earlier than 90 days after the 2026-09-07 notice sent to every organization with active credentials — so not before 2026-12-06 — and only in a release after the rename itself. If you maintain dashboards or scheduled queries against these tables, plan to move them to the new names during that window.