Every track, ordered by whether it needs you rather than by how much it
matters. The bar is the shape of the work: ready, waiting on a ruling,
blocked, parked. Read the colour, not the numbers.
State and counts are DERIVED from issue labels on every build — no word
on this page is one somebody has to remember to update. The impact line under each track
is the exception, and lives in
docs/status/track-order.json.
Generated 2026-10-14 05:59 UTC.
p0 work is startable now — answer-truth.
922open
871ready to start
0waiting on you
3blocked
48parked
These four are exclusive and sum to 922 — every
open issue is counted once. 1 issues carry both state:parked and blocked; they count as blocked here, so the parked tile reads lower than the label's own count. Blocked wins because it is the one you cannot clear by deciding.
Next: [answer-truth] EPIC: answer questions exactly and completely over… #4266
An aggregation question (spend, count, list-every) gets an answer summed from a top-k retrieval sample instead of the full matching set, and states it with confidence.
Next: [L1-g2-w10] [ama-disclose] enterprise messages of 10 characters or… #4652
Content reaches someone who should not see it. Nobody experiences this until it is catastrophic, which is why it loses to everything above and needs a standing allocation instead.
Regina's people are in the graph under the wrong names, and 40% of her mail is held out of embeddings, extraction and search by a flag no one can explain per-row. Attribution is repaired in production; visibility is not.
Next: [fold-efficiency] the fold's dominant cost is graph_cross_conv_matching… #3495
Regina's graph was wrongly merged and ingestion frozen for 9 days; the repair is done and ingestion resumed 2026-08-25 — the residual risk is un-verified repair side-effects (stale embeddings, the unsigned OCR pass) and the fold-cost class that caused it.
Next: [deploy-safety] a status-board label error turns a green production… #3946
A deploy or shared machine silently degrades the running product — the serving tier starved at 7.54% CPU steal until 2026-08-23; the remaining item is finishing the poll cutover so the old cron driver is retired.
Next: [fold-efficiency] the fold's dominant cost is graph_cross_conv_matching… #3495
The cross-source fold exceeds its 3h ceiling without document extraction's help, so new mail goes unscoped when the fold dies before scope_classification — and the doc-needs precompute it feeds is 87% dead weight by rows.
24 startable
ama-disclose-audienceunranked5 open
moving
Next: Node text and conversation_* spec text drift apart across re-extraction… #3387
5 startable
ama-disclose-org-membershipunranked3 open
moving
Next: [ama-disclose] disclosure gate keyed on org_membership —… #3596
2 startable · 1 parked
ama-trust:spreadsheetsunranked6 open
moving
Next: [ama-trust:spreadsheets] AMA refuses to aggregate a column that IS… #3996
5 startable · 1 parked
arabic-retrievalunranked32 open
moving
Next: [arabic-retrieval] cost the tiered-extraction option before widening… #4270
28 startable · 4 parked
ask-canaryunranked3 open
moving
Next: [ask] Per-screen page title and TTI marks belong at the route, not in… #4376
3 startable
client-authunranked2 open
moving
Next: Web Google signup can never complete — the OAuth full-page redirect… #2768
2 startable
conversation-firstunranked7 open
moving
Next: [Tracker] Per-slice Materializer.add_intent() paths for direct repo… #1341
6 startable · 1 blocked
conversation-truthunranked3 open
moving
Next: [conversation-truth] Should conversation membership be a SEPARATE graph… #3904
3 startable
identity-dedup-chokepointunranked4 open
moving
Next: No choke-point records that two identities are the same human — 10… #3416
4 startable
person-name-truth:low-valueunranked7 open
moving
Next: [L1-g2-w9] [person-name-truth] is_low_value is a one-way ratchet — a… #3855
6 startable · 1 parked
viewer-account-resolutionunranked7 open
moving
Next: VIEWER-TRUTH: the viewer's display NAME has 5-6 implementations that… #3173