The Version table prints the version's last day as its End Date. Read as
an ordinary day-precision end, the event left the calendar 9 to 24 hours
before the pass did. The owner's timer on Europe put the end at about
20:00 UTC on 29 Oct: Asia's 04:00 reset opening the 30th, for every
server. The source's entry now says so, with that evidence, and the
event ends at exactly 2026-10-29T20:00Z. Found in the whole-branch
review.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Records why Miliastra was reverted and why Endfield's Protocol Pass has
no source yet, updates AGENTS.md's parser list and marks sub-project 3
in TODO with the follow-ups.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Fandom's Event page, an existing source, already lists the pass as
"Silvery Moonlight Chronicle" with the same dates, and the merge folded
the new copy into it. The feed was unchanged, so a second source for one
event is cost with nothing to show. This reverts 9f8150f.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
New `fandomtable` parser (src/ingest/parsers/fandomTable.ts) reads Fandom's
dated pass-history table by its Name/Start/End headings and emits runs not
yet ended by ctx.now, at day precision. Titles come from a per-source
PASS_SOURCES table so the next pass pages are one entry each.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Both fortnightly. Beyond the Rails is ticked once a cycle and rolls on
the Asia server's reset for every reader: on Europe the owner read it
ending Wednesday 21:00 UTC, Asia's Thursday 05:00. The Fund counts fons
against the guide's base cap of 1,000,000; the note says a reader's own
cap is higher and how to set it.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The Hazard Zone on a 28-day cycle, ticked once, anchored on the reset the
owner read in game (Monday 2026-11-09). The source says "roughly", and
the note says so.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Both every two weeks on alternating Fridays, ticked once a cycle,
anchored on the owner's in-game readings: Shiyu rolled 2026-10-16,
Deadly Assault rolls a week later.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Pure Fiction, Memory of Chaos and Apocalyptic Shadow on their six-week
rotation, two weeks apart, each ticked once a cycle and anchored on the
owner's in-game end dates; and the Divergent Universe / Currency Wars
pool, weekly since 4.2, at the owner's 18,000.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The spec and plan say what shipped and what is still open. TODO.md gains the
next two sub-projects with the facts already gathered, the design questions to
settle and the sources to check, and a default check-in goal for To Temper
Thyself and Journey Far.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Game-Time-Master lists NTE Europe on UTC+1, but the game disagrees: on
2026-10-16 two in-game timers ran out at 03:00 UTC — one on 26 Oct, after the
clocks change — and Realm of Greed says "resets every Monday at 5:00". Only a
fixed UTC+2 fits all three.
Re-labels the hour 03:00-04:00 UTC of every European reader's NTE day.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The design spec (entries with the sentence each rests on, what was looked for
and not shipped, and the facts gathered for N-week cycles and dated pass
windows), the plan, PRD F12 and F17, DATA-MODEL's prefs, and TODO.md's open
in-game checks.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Game8 and Fandom name Genshin's anniversary selectors with no word in common, and both run the same window, so the calendar showed each selector twice. Game8's own blurb names the Fandom event ("…limited 5-star selector, Tabletop Troupe: A Gathering on Adventure's Eve!"), which is how a person paired them, so merge now reads it the same way.
The rule is narrow on purpose: the whole title of two or more words, both boundaries within 24h, and never for a copy that already has a twin by title in the other source, since then the blurb is naming a neighbour such as a festival's shop. Measured on every two-source page held, it fires on the two selector pairs and otherwise only where titles already matched. Dates alone never pair: patch day puts about twenty Genshin events on one window. aliases.ts is the recorded-pair fallback for copies no page links; it is empty.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
titleExtends caught a source appending words ("Bedazzling Dawnstar Sign-In") but not prepending them, which is the shape Game8's "Battle Pass - …" label took. The same qualifier at the other end now pairs too. The two-word floor holds at both ends, because a one-word base would fuse half of a patch day.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
When merge fuses two sources' copies, the loser's id leaves the calendar, and it may already hold a reader's ticks from a day it was published alone. Until now that needed a hand-edited RETIRED_IDS entry per fusion.
The survivor now lists what it absorbed in an optional mergedIds field. When a feed arrives the client turns those into carry-forward pairs, skipping any id the feed still publishes, and an import takes the same table. Optional, so an older cached feed still validates and an older bundle drops it unread. RETIRED_IDS stays for parser changes, whose old ids no feed will name again.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Game8 heads Genshin's pass "Battle Pass - Moontrace"; Fandom lists "Moontrace" and files the category in a column. The label minted a second id for one event, so every battle pass since the Fandom source landed has been on the calendar twice, and a reader reported it.
Both sources now mint the same id and merge fuses them by id. The two Game8 ids that moved, Frostfarer and Moontrace (the only ones git history shows ever shipped), go in RETIRED_IDS so readers' marks follow.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
An event id is a localStorage key, so a fix that changes one orphans every tick, note and ignore filed under it. retiredIds.ts holds an explicit table of old id -> new id, and the progress, daily and ignored stores move entries across on load and on import.
It moves rather than copies: a copy left under the old id would be carried forward again on every load and bring back a mark the reader had cleared. When both ids hold an entry, each store's own import merge decides, except that progress keeps any field only the older row held. The table starts empty; the change that moves ids adds its entries.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
PRD F13 said the form asked how often after Kind and had no word on
examples; F17 and DESIGN.md had no example row. Adds the spec and plan,
and ticks the TODO item.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
SOURCES.md § 16 holds the diagnosis (run 317's rejected zeros, the page
unchanged in shape with nothing current since 19 Sep) and the replacements
assessed and declined: NamuWiki behind a Cloudflare challenge, an official
site with no event posts, no Game8 page. AGENTS.md says when a person may set
`lapsed` and what it does not excuse. TODO ticks the item and notes the
cosmetic "last confirmed never" seen in the CI state.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The data model, PRD F17, README and DESIGN described one global chores
switch and a checkbox per game for built-ins. Adds the spec and plan, and
ticks the TODO item.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Giving Genshin's "To Temper Thyself and Journey Far" its five-of-seven goal
meant finding it in the list, opening its sheet, saying it repeats and only
then setting the goal. The strip now has an Edit mode beside Catch up that
lists its events by game with their goals, opens the sheet's goal control in
place, and adds any running event a game at a time. It calls the same
setDaily/setGoal the sheet does, so it can store nothing the sheet would not.
Weekly and monthly tasks are not due tonight, so they move to a row under the
dailies, and each period can be switched off (showWeeklyTasks,
showMonthlyTasks; default on, so nothing disappears on upgrade).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A new reader should see what's coming before deciding they don't want
it. It's set at first-run setup rather than in the defaults, so an
existing install whose saved prefs predate the field keeps its answer
instead of having it switched on underneath it.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Four lessons were known to one agent's private notes and to nobody else.
They now live where every agent reads: how Gitea Actions differs from
GitHub Actions, the owner's decision to refresh unreachable sources by
hand rather than relax the robots gate, how to look at the built app in
a headless browser and its traps, and one command per shell call.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
When the network failed behind the service worker, it answered from its
own store. Check for new data then said nothing newer was published when
the server was never reached. The worker now marks that copy, and a fresh
check treats it as unreachable; ordinary loads still use it offline.
The built-in tasks entry linked to a follow-up fix instead of the
feature, and the changelog spec's details now match what shipped.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The changelog's own entry names the commit before this one, since a
commit cannot contain its own hash. AGENTS.md now asks every
reader-visible change for an entry, and every multi-commit feature for a
note, so the record keeps up. PRD F18 and the storage and design records
describe what shipped.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Make it mine's copying was inline in App and untested; it is now a pure
function with a test, so a change to either store can't make the copy
silently arrive empty. The question's buttons now say what they answer
when heard out of context. The prefs block and the spec now match what
shipped.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A reload can be served the browser's cached feed for minutes. The button
beside the data's age asks the server past that copy, swaps in a newer
feed without reloading, and says honestly what came back. Newer is
judged on events and sources, never the build stamp. The age now says
"just now" rather than "ended ago" when a fresh feed is a moment ahead
of the page's clock.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The changelog has two tiers, each tied to git: short entries shown after
a reload, and release notes summarised from the history and specs.
Ridu Weekly (1,300) and the Weekly Routine (10) are now sourced. The
first guesses of 2,100 and 10 were checked against the owner's links.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Review found that import and edit rebuilt a task from this build's parse.
That stripped any field it didn't know, and import dropped a newer
version's task entirely. Both now keep the record as stored and validate
only to display it.
An event whose goal is points can't be ticked, so it no longer counts as
waiting today: it counts with its week. The month-day field reset
itself to 1 when emptied; it now keeps a draft.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The product, storage and design records now cover F17. They say why the
task and amount stores, unlike the custom-event store, never filter on
load, which is the rule a future store would most likely copy wrongly.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Task status reuses the goal status shape, so one chip, one count and one
tally serve chores, events and tasks alike.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Review found three ordinary actions that deleted a goal this build could
not read: "It repeats daily", pressing an already-pressed Off, and "not
daily" without saying so. Only "not daily" clears a goal now, and its
warning keys off what is stored, not what parses.
The count field was bound straight to the goal, so emptying it snapped
the old number back and typing appended to it. It now keeps a draft.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The product, storage and design records now say what a goal is, why it
never marks an event done, and why an unreadable goal is kept rather than
stripped. That last rule is the one a future cleanup would most likely
undo by mistake.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The goal control lives on the checklist only: "It repeats daily" already
reaches it from a non-daily event, so a second entry point added nothing.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The goals draft stopped at three open questions; the owner answered them,
and widened the scope to weekly and monthly chores with numeric targets.
Both share one period engine, so they are specified together and built
goals-first.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Rewrites the reset table for every corrected game, records the rule that a
game's offset must make its day key name the source's calendar day, and
names Game-Time-Master as the evidence for server clocks a source does not
state (transcribed, never vendored: it is GPL-3.0).
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A game's servers need not agree on the reset hour any more than on the
offset. Reverse: 1999 and Nikke spell their 05:00 out for all three
regions, so no reader's day key moves.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
