Observed
Play 2026-08-27 afternoon, look-only http://127.0.0.1:4220/ (already-running tab). Sidebar walk only.
- Discover hero: Proposals 0 · waiting in the evidence pipeline
- Findings inbox hero: Proposals 0 · Pi artifacts available
- Same app, Findings ad-hoc panel: Proposal queue · Unreviewed hypotheses · 1 waiting with real rows
A first-timer trusts the zeros and never opens the drawer.
Adjacent to #101 (1 waiting vs 4 unreviewed rows inside the queue). The new part is the 0 on the two front-door tiles, not the 1-vs-4 spread. Research cases Open 4 / 4 retained vs Review Tracked cases 1 is the #95 family — do not refile #95.
Why it matters
The two doors an operator sees first say nothing is waiting. The only queue that has work is below the fold / in another panel.
Proposed direction (not a spec)
Front-door tiles should count the same unreviewed hypotheses the Proposal queue shows, or say they count a different thing (Pi artifacts only). Copy/count only. Do not add a spend path.
Out of scope
Observed
Play 2026-08-27 afternoon, look-only
http://127.0.0.1:4220/(already-running tab). Sidebar walk only.A first-timer trusts the zeros and never opens the drawer.
Adjacent to #101 (1 waiting vs 4 unreviewed rows inside the queue). The new part is the 0 on the two front-door tiles, not the 1-vs-4 spread. Research cases Open 4 / 4 retained vs Review Tracked cases 1 is the #95 family — do not refile #95.
Why it matters
The two doors an operator sees first say nothing is waiting. The only queue that has work is below the fold / in another panel.
Proposed direction (not a spec)
Front-door tiles should count the same unreviewed hypotheses the Proposal queue shows, or say they count a different thing (Pi artifacts only). Copy/count only. Do not add a spend path.
Out of scope