Release notes
Changelog
Everything shipped in Relay Fan-Out so far. You're on v1.0.2.
v1.0.2
Current- Fixed invite workflow: added a proper unique email constraint so invitations save and re-invites update correctly.
- Version bump and release housekeeping.
v1.0.1
Aug 1, 2026- Signed-in users now have every fan-out saved to their history automatically.
- History entries can be deleted from the History page at any time.
- General stability and release housekeeping.
v1.0.0
Jul 31, 2026- Live progress reporting across every stage: query expansion, site search, per-page SEO scans with counts, and SEO plan build.
- Per-sub-query and per-site status badges so you can see what's still being analyzed.
- Custom fan-out node logo, favicon, and installable PWA support (add to home screen, standalone display).
- Renamed the app to Relay Fan-Out across the header, page titles, and metadata.
- Design polish: Inter for headings and body copy, JetBrains Mono reserved for data labels, refined search surface and hero.
v0.9.0
Invite-only access- Public registration disabled — access is invite-only.
- Admin Invites page to send and revoke invitations by email.
- Dedicated /join registration flow that only allows sign-up when the email matches an invite.
- Google sign-in added alongside email and password.
v0.8.0
SEO intelligence- Comprehensive SEO plan generated at the top of the page from every ranking page's factors and gaps.
- Signed-in users get SEO details pre-loaded in the background so they're ready on arrival.
- Signed-out users see the query fan-out only; sites and SEO analysis require an account.
- Expandable chevron on each cited site with an AI breakdown of why it ranks: title and meta, content, structure, schema, authority, freshness, plus gaps to exploit.
v0.7.0
Live citations- Real web search per sub-query, showing the actual websites currently cited.
- Clickable, domain-labeled source lists under each sub-query.
- Dark and light mode toggle with saved preference.
v0.5.0
First release- Real AI query fan-out: expand any query into the sub-queries an AI engine would actually run.
- Sub-queries grouped by intent with a rationale and likelihood score for each.
- Accounts with saved analysis history.
- Dark, console-inspired design system.