Loot & Baloot

v3.2.7 (2026-05-14)

Full Changelog Previous Releases

• docs(CHANGELOG): add v3.2.7 release notes
  v3.2.7 entry covers:
  • TAKWEESH tooltip now explicitly states only opposing-team
    illegal plays qualify (closes the v3.2.6 investigation's
    Scenario B UX hazard).
  • Tooltip names both qualifying proof patterns (false-AKA
    marker, later same-suit reveal after off-suit).
  • Same-team Takweesh runtime behaviour unchanged — purely
    tooltip wording change.
  • No gameplay logic, protocol, saved-variable, scoring, .toc,
    .pkgmeta, .github, or packaging changes.
    Verification: harness 1297/0 (+2 BN source-pins from v3.2.6
    baseline of 1295/0), H1=11/0, H7=9/0.
    Committed locally only — no push, no tag, no release until
    Codex reviews.
    Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com
• ui(UI.lua): clarify TAKWEESH tooltip same-team rule + qualifying patterns
  v3.2.7 UX-only follow-up to the v3.2.6 AKA/Takweesh fix.
  Tooltip wording update at UI.lua:2491-2508 (the only TAKWEESH
  tooltip call site):
  Pre-fix: tooltip described the off-suit/revoke realism gate
  pattern only ("the violator later showed they had the led
  suit"). Silent on the false-AKA carve-out shipped in v3.2.6
  and silent on the same-team-Takweesh wrong-call rule.
  Post-fix: tooltip now (a) names both qualifying patterns —
  "false AKA marker or a later same-suit reveal after an
  off-suit play" — and (b) explicitly warns "Only OPPOSING-
  team illegal plays qualify; calling Takweesh on your own
  teammate counts as a wrong call." Closes the v3.2.6
  investigation's Scenario B UX hazard.
  Two phrases are deliberately kept on single source lines so
  the BN.1 / BN.2 source-pin Lua-pattern finds anchor without
  being defeated by line-concatenation breaks.
  Tests: new BN section in tests/test_state_bot.lua locks both
  phrases. BN.1 anchors the same-team rule; BN.2 anchors the
  qualifying-patterns parenthetical. The underlying same-team
  filter behavior is already wire-locked by BM.3 — no new
  behavioral test needed.
  Harness: 1295 → 1297 (+2 source-pin asserts).
  H1/H7 smokes clean.
  No gameplay, runtime decision, protocol, saved-variable,
  scoring, .toc, .pkgmeta, .github, or packaging change. The
  rendered tooltip text is identical to the user-approved
  wording in the Codex implementation prompt; line-flow inside
  UI.lua's string-concatenation block was adjusted only to make
  the source-pin anchors land on single lines.
  Design doc: .swarm_findings/v3_2_7_takweesh_tooltip_design.md
  Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com
• docs: add v3.2.7 TAKWEESH tooltip clarification design
  UX-only design pass for the v3.2.6 follow-up flagged in
  v3_2_6_aka_takweesh_investigation.md §9.3. Identifies the
  TAKWEESH button tooltip at UI.lua:2476-2485 as the user's only
  pre-confirmation safeguard against the same-team Takweesh UX
  hazard (caller's team penalized when they Takweesh their own
  teammate's false-AKA marker).
  Proposes a narrow text-only change to the tooltip wording
  adding the "Only OPPOSING-team illegal plays qualify" rule.
  Single source-pin source assert recommended (BN.1 after the
  v3.2.6 BM section); the underlying same-team filter behaviour
  is already wire-locked by BM.3.
  No gameplay/protocol/saved-variable/scoring/.toc/packaging
  change. Justifies a small v3.2.7 patch tag.
  Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com