Saved and read history
Use Later for articles to revisit, Starred for references, and Archive for articles tucked out of normal views. Done records that you have finished reviewing an article; opening an article alone does not mark it Done. A star is independent of workflow state.
Find an article again
| Where | What to find |
|---|---|
| Later | Articles set aside for future reading. |
| Starred | Articles you have starred. |
| Archive | Archived articles, still searchable. |
| Search → State → Done | Articles you marked Done. |
Use Search filters and Saved Views to combine state, source, category, date, tag, or starred status. Include archived articles when looking beyond active views. You can reopen a result and change its triage state or star again.
Reading DNA and organization
Reading DNA summarizes your reading mix and recommendation controls, including category/source activity, streaks, goals, and related learning tools. Your interactions are scoped to your account. See Recommendations for how your actions affect ranking, and Organize and learn for Collections and other ways to retain useful articles.
Export and restore
In Settings → Data Export, choose Download archive to save a JSON archive of your reading history, article metadata and workflow state, briefings, source subscriptions, and preferences. Cached article body text and secrets such as passwords, session tokens, and push keys are excluded.
Choose Restore archive to select a previously downloaded JSON file. Existing articles, briefings, AI memories, subscriptions, and preferences are matched and updated instead of duplicated. The result reports added, updated, and skipped counts. Restore writes only to your current account and requires the archive schema version supported by the running server.
Source restoration covers enabled/disabled shared subscriptions and private sources owned by your account. It does not take over another user's private source. Preferences include recommendation weights, onboarding choices, and notification settings; secret credentials and push subscriptions must be set up separately. See Settings and account data for the full data-management workflow.
Retention and troubleshooting
Snoozed articles can return when their timer expires. Analytics events are
pruned according to ANALYTICS_RETENTION_DAYS (default 180 days); that is
separate from deleting your article history.
If history appears missing, verify your account and Search filters before restoring a backup. If restore fails, read the displayed error and use an archive with the schema version supported by the running server. Keep downloaded archives private: they contain personal reading activity even though secrets are excluded.