0.23.0 adds the LeetCode review top-up earner (20 shutdown minutes for a
finished review, in the registry from 2026-11-23). Pin-only for this repo,
plus the one test that assumed the stand-in earner is last.
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWXYyzSw5S9UTM5k5M1
The server serves the git-ignored web/dist, so a deploy that only pulled
and restarted kept the last hand-built bundle live; the kuhy credit footer
never reached :8770. deploy.sh now rebuilds it when the web/ tree differs
from the last built one (--web-only runs just that step).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The shared kuhy_credit logo (kuhyx/utils, kuhy-credit-v0.3.1) now sits
bottom-right on every lock surface, the aux windows, the status window
and the heat-skip prompt; at the end of the workout app's Settings;
and in the web dashboard footer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The old break rule ran at finish after the session was saved, so
daysSince was always 0 and it never fired. Now the gap deload cuts
by tier (7-13d 1 step, 14-20d 2, +1/week, 40% cap) when the workout
loads, stamps last_deloaded_at (shared with manual deload, synced),
logs progression_events, and shows the old target struck through.
One-time backfill with pre-cut export + restore script.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The watch was armed in __init__, before the lock window build
(xrandr, setxkbmap, surfaces, grab), so the ~1.5 s build was reported
as a tk.mainloop freeze. Arm it via after_idle in run() instead; a real
freeze after the loop starts is still reported.
Error-Signature: screen-locker:c890f31a21
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
deploy.sh records the workout app's git tree after each phone install
and skips the phone when the tree is the same, so a PC-only change
deploys with the phone unplugged. No record means the phone is needed.
Todo: 2556952d
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
pnpm build now writes the short HEAD hash to web/dist/version.txt
(web/stamp-version.mjs). The web server serves it at GET /version as
text/plain and answers 404 when the stamp is missing, so an unstamped
build never passes a live-build check. Docstring now says Astro.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
gatelock error_report: uncaught, thread and Tk callback errors plus a
frozen lock window over 1 s go to the app-errors spool. The pre-commit
pylint env moves to gatelock 0.16.0, the version requirements pin.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
kuhy, 2026-11-18: shutdown 18:00, a run ending 19:58 must give 20:00
and the book's +1h 21:00 -- what the daily reset derives. The
late-credit rule (credit's own time + reward, 2026-10-17) made it
21:58 and 22:58. Unpaid gates now hold the PC behind the evening
chain until 23:00 instead, so the rule and its credit-time plumbing
go: _late_credit, the at= anchors, _done_at. The preview flags any
live-vs-derived MISMATCH again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWWLoUxxSPf1m8KUVo8
A late credit moves shutdown from an hour already past to the run's
end plus its reward, so the toast read "+3h58m" for a 2h run
(18:00 -> 21:58, 2026-11-18). The delta is now capped at the reward
the credit earned: still the real move when the ceiling cut it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWWLoUxxSPf1m8KUVo8
Pin earned-time-v0.22.0. The toast and the poke reply show only the
shutdown gain; gaming_budget_minutes leaves the reply and the phone
app's result line (an older PC's field is ignored).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWViGJJaqBKGr1Hbppr
wait_for_turn sleeps its 1.5 s startup grace for real, so every test
that builds a locker paid it. The suite took ~135 s under the CPU
cap and the ci-pytest gate step could not finish inside the 60 s
action cap. With the grace zeroed in the hermetic gatelock fixture
all 2142 tests run in 13 s.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWWfb63jncj9Lca2jeX
On 2026-11-16 a manual workout was credited at 07:59 by
_session_stream, but the open lock window never rechecked log.json
(TRY AGAIN only asked StrongLifts/adb), so kuhy had to mark a sick
day. The window now polls log.json mtime every 3 s via Tk after and
stands down (decision workout_credited_elsewhere, no second credit),
and TRY AGAIN stands down when today is already logged.
_verify_phone_workout is deliberately unchanged (shared with
early-bird/status).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWWfb63jncj9Lca2jeX
Revert 6777656; the app again writes zero distance across a gap
over 30 s. Also drop the importer-test skip for its fixture.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWYwyP67ppvkeie3X1F
Every locker screen, the status window and the credit toast now show
glyphs (screen_locker/_glyphs.py) and numbers, coloured by state.
Sentences that carried a reason (verifier message, sync diagnosis,
sick refusal, form rejection) go to the log at warning instead.
New pre-commit hook text-free runs scripts/verify_text_free.py.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
textgate scan 319 -> 0: docs/explanations (dropped explain-diff
output, not published, nothing links it), Tk window titles, the
MaterialApp title, '+kg' -> '+' plus the weight icon, and
spoken-only Semantics labels hoisted so the gate sees them.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Painted chrome becomes icons or symbols; old words stay as
spoken labels or code comments. OS window titles kept. Tests
build today from the local day.
Todo: ed07cff3
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The PC keeps a dropout moving at 1 m/s to 25 km/h; the app zeroed
every gap over 30 s. Bridge unpaused dropouts, pauses stay zero.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWXcjiPpgRfxzrRfoky
Silent-failure checker, mypy, pylint now clean. Hidden packages count
as installed (pm list -u), pull gets a long timeout, wifi entries are
matched by ro.serialno, and a missing log/drop dir is a warning.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWZ9CKZsNxtD5Jr9bYD
Anchor = PC credit time of an app-authored run (joined on TCX end
time in log.json), due 7 days later, gated on a validated import
marker, then adb backup + uninstall. Typed results, --status/--dry-run.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWZ9CKZsNxtD5Jr9bYD
Pure-Dart TCX codec pinned to the PC's RunnerUp parser by a shared
golden fixture (Dart byte-compare + pytest through the real verifier),
GPS filters and moving-time maths mirroring the PC walk rule, runs /
run_points tables (schema v9), PositionSource seam over geolocator,
location permissions and foreground service type, map/cache deps.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWZ9CKZsNxtD5Jr9bYD
Poll morning_session every 15s (05:00-11:00) while the lock is
up; on a signed skip bank the early-bird marker, record
wake_alarm_skip_late and leave through the normal close().
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWZd3smuxDGmEgXkYP8
The Firebase stream only ingested StrongLifts sessions, so a manual
workout logged on the phone waited for the 15-min workout-sync tick.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWVTJPDdbeF71W9Wd3q
Real one-line docstrings stating what each test pins down. Files that
would pass 250 lines moved whole classes into _part2 siblings
(sick_panel, earned, poke_server, shutdown_reading,
manual_workout_dialog).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWYdwRAtMg7ki6TKxuU
Real fixes: split over-long functions (lock-decision chain, manual
sync, validation, phone verification, status), hoist a dead lazy
import, context-manage Xvfb and temp dirs, add encodings, share the
silent-failure argv loop.
lock_grab_handoff read lock._recovery, which gatelock's LockWindow
does not have (its public property is recovery); a contract test now
checks every attribute the handoff reads against the real class.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WzxdWYdwRAtMg7ki6TKxuU