SeymourBETA

Update 2.60

6/9/2026

v2.60.0BETA

Seymour

Reusable read-only Atlas evidence badges, applied to the Monster AI Editor

First layer of HANDOFF_ATLAS_READONLY_MODULE_BADGES: a reusable evidence-badge component that shows, per entry, where the data came from and how far it can be trusted — with no writer, no apply, no SIN. New Modules/Common/AtlasEvidenceBadgeStrip.axaml(.cs) (drop-in UserControl with an Evidence property) + Modules/Common/AtlasEvidenceBadge.cs (AtlasEvidenceBadge colored by severity + AtlasEvidenceInfo carrying badges/origin/"where it appears", factories ForBibleEntry/ForDetail). Key decision (after the adversarial review): ONE engine — AiBibleEvidence was parameterized into Derive(...) and the strip reuses the entry's canonical badges via entry.EvidenceBadges, so the Monster AI inline strip and the full BIBLE window can never disagree; IDA stays a separate structural badge and never collapses into a green proved (a bug the review caught and fixed). Added the mission's partial token. The "where it appears in game" text became AiBibleWhereAppears (one DRY source shared by BIBLE and the modules). Applied to the Monster AI Editor inline BIBLE panel: badges + origin + a "where it appears" expander for the selected entry; the existing 📖 Abrir guia completo button opens the focused BIBLE. No runtime, SIN, save-path, or SpiraDataAtlasCatalog provider changes. Build 0 errors / 361 warnings; --spiraatlas-rt0 and --aicmdmeta-rt0 PASS; AiScriptLab --ai2 PASS (validator clean 346/346)

v2.59.0

v2.60.1BETA

Seymour

Atlas evidence badges in the Sphere Grid Explorer (second read-only application)

The v2.60.0 badge infra got its second home: selecting a node in the Sphere Grid Explorer now shows the node's evidence badges (parser-corpus + RT0 + read-only), origin, and "where it appears", via AtlasEvidenceInfo.ForDetail + <common:AtlasEvidenceBadgeStrip>. A tiny read-only accessor SpiraDataAtlasCatalog.TryGetSphereGridNode(layout, nodeIndex) (mirrors TryGetMonsterDetail) maps SourceKind (Original/Standard/Expert → OSG/SSG/ESG) + node index to atlas:sphere-grid-node:{layout}:{index} — the exact id the catalog builds, so the node resolves; on any miss the strip hides itself (no false badge). Crosslink honesty (guardrail re-stressed by Halyson): of the 4 requested modules, only Sphere Grid has a stable crosslink. Treasure (opaque sourceKey, not reverse-derivable from the buki_get index), Shop (the atlas key is bank+slot and item_operand_hex is empty for items; mapping by operand would be a category error), and Mix (no Mix detail kind exists in the atlas) were intentionally left out — forcing a badge there would break the "no stable crosslink → hide" guardrail. No writer, SIN, runtime, or badge-engine changes. Build 0 errors / 361 warnings; --spiraatlas-rt0 PASS (2493 sphere-grid nodes) and --aicmdmeta-rt0 PASS. Adversarially verified: 0 defects

v2.60.0

v2.60.2BETA

Seymour

Atlas evidence badge in the Treasure Editor (gear chests) + honest correction of the Treasure crosslink

The HANDOFF_ATLAS_RIKKU_TREASURE_SHOP_MIX_CROSSLINK_SCOUT scout reopened the v2.60.1 "Treasure has no stable key" claim and proved the opposite for gear chests: the treasure-buki-get crosslink source_key is NOT opaque — it is treasure:0x{index:X4}, reconstructible straight from the editor's row index. Byte-grounded check: the crosslink's (treasure_index, buki_get_id) ≡ takara.bin's (logical_index, type) across all 82 Kind=0x05 rows, 0 mismatches. New read-only accessor SpiraDataAtlasCatalog.TryGetTreasureGear(treasureIndex, expectedBukiGetRow) resolves atlas:gear:treasure-0x{index} with a value-guard: since Treasure is an editable module, the badge only shows while the chest's current buki_get still matches the corpus — repointing it or changing the Kind hides the strip (no stale badge). Wired into the TreasureEditor "Gear Pickup / buki_get" card via AtlasEvidenceInfo.ForDetail + <common:AtlasEvidenceBadgeStrip> (badges parser-corpus+RT0+read-only, never proved). Shop stays partial (the positional 0x{bank}:slot:{n} key IS stable and reconstructible, but item-shop only carries display_text and gear-shop carries no per-slot index, so there is no clean value-guard on an editable module → UI deferred to avoid a false positive). Mix stays blocked (no Mix detail kind exists in the atlas; adding one would invent an anchor). BukiGet = redundant/1:many (already self-references takara). The --spiraatlas-rt0 gate gained 3 durable asserts (positive lookup + value-guard + non-gear miss), all PASS. No writer, SIN, runtime, or Common badge-engine changes. Build 0 errors / 361 warnings; --spiraatlas-rt0, --aicmdmeta-rt0, and --treasure-rt0 PASS. Doc: docs/ai/ATLAS_RIKKU_TREASURE_SHOP_MIX_CROSSLINK_SCOUT_RESULT_2026-06-09.md

v2.60.1

Back to updates

SeymourBETA

Update 2.60

6/9/2026

v2.60.0BETA

Seymour

Reusable read-only Atlas evidence badges, applied to the Monster AI Editor

First layer of HANDOFF_ATLAS_READONLY_MODULE_BADGES: a reusable evidence-badge component that shows, per entry, where the data came from and how far it can be trusted — with no writer, no apply, no SIN. New Modules/Common/AtlasEvidenceBadgeStrip.axaml(.cs) (drop-in UserControl with an Evidence property) + Modules/Common/AtlasEvidenceBadge.cs (AtlasEvidenceBadge colored by severity + AtlasEvidenceInfo carrying badges/origin/"where it appears", factories ForBibleEntry/ForDetail). Key decision (after the adversarial review): ONE engine — AiBibleEvidence was parameterized into Derive(...) and the strip reuses the entry's canonical badges via entry.EvidenceBadges, so the Monster AI inline strip and the full BIBLE window can never disagree; IDA stays a separate structural badge and never collapses into a green proved (a bug the review caught and fixed). Added the mission's partial token. The "where it appears in game" text became AiBibleWhereAppears (one DRY source shared by BIBLE and the modules). Applied to the Monster AI Editor inline BIBLE panel: badges + origin + a "where it appears" expander for the selected entry; the existing 📖 Abrir guia completo button opens the focused BIBLE. No runtime, SIN, save-path, or SpiraDataAtlasCatalog provider changes. Build 0 errors / 361 warnings; --spiraatlas-rt0 and --aicmdmeta-rt0 PASS; AiScriptLab --ai2 PASS (validator clean 346/346)

v2.59.0

v2.60.1BETA

Seymour

Atlas evidence badges in the Sphere Grid Explorer (second read-only application)

The v2.60.0 badge infra got its second home: selecting a node in the Sphere Grid Explorer now shows the node's evidence badges (parser-corpus + RT0 + read-only), origin, and "where it appears", via AtlasEvidenceInfo.ForDetail + <common:AtlasEvidenceBadgeStrip>. A tiny read-only accessor SpiraDataAtlasCatalog.TryGetSphereGridNode(layout, nodeIndex) (mirrors TryGetMonsterDetail) maps SourceKind (Original/Standard/Expert → OSG/SSG/ESG) + node index to atlas:sphere-grid-node:{layout}:{index} — the exact id the catalog builds, so the node resolves; on any miss the strip hides itself (no false badge). Crosslink honesty (guardrail re-stressed by Halyson): of the 4 requested modules, only Sphere Grid has a stable crosslink. Treasure (opaque sourceKey, not reverse-derivable from the buki_get index), Shop (the atlas key is bank+slot and item_operand_hex is empty for items; mapping by operand would be a category error), and Mix (no Mix detail kind exists in the atlas) were intentionally left out — forcing a badge there would break the "no stable crosslink → hide" guardrail. No writer, SIN, runtime, or badge-engine changes. Build 0 errors / 361 warnings; --spiraatlas-rt0 PASS (2493 sphere-grid nodes) and --aicmdmeta-rt0 PASS. Adversarially verified: 0 defects

v2.60.0

v2.60.2BETA

Seymour

Atlas evidence badge in the Treasure Editor (gear chests) + honest correction of the Treasure crosslink

The HANDOFF_ATLAS_RIKKU_TREASURE_SHOP_MIX_CROSSLINK_SCOUT scout reopened the v2.60.1 "Treasure has no stable key" claim and proved the opposite for gear chests: the treasure-buki-get crosslink source_key is NOT opaque — it is treasure:0x{index:X4}, reconstructible straight from the editor's row index. Byte-grounded check: the crosslink's (treasure_index, buki_get_id) ≡ takara.bin's (logical_index, type) across all 82 Kind=0x05 rows, 0 mismatches. New read-only accessor SpiraDataAtlasCatalog.TryGetTreasureGear(treasureIndex, expectedBukiGetRow) resolves atlas:gear:treasure-0x{index} with a value-guard: since Treasure is an editable module, the badge only shows while the chest's current buki_get still matches the corpus — repointing it or changing the Kind hides the strip (no stale badge). Wired into the TreasureEditor "Gear Pickup / buki_get" card via AtlasEvidenceInfo.ForDetail + <common:AtlasEvidenceBadgeStrip> (badges parser-corpus+RT0+read-only, never proved). Shop stays partial (the positional 0x{bank}:slot:{n} key IS stable and reconstructible, but item-shop only carries display_text and gear-shop carries no per-slot index, so there is no clean value-guard on an editable module → UI deferred to avoid a false positive). Mix stays blocked (no Mix detail kind exists in the atlas; adding one would invent an anchor). BukiGet = redundant/1:many (already self-references takara). The --spiraatlas-rt0 gate gained 3 durable asserts (positive lookup + value-guard + non-gear miss), all PASS. No writer, SIN, runtime, or Common badge-engine changes. Build 0 errors / 361 warnings; --spiraatlas-rt0, --aicmdmeta-rt0, and --treasure-rt0 PASS. Doc: docs/ai/ATLAS_RIKKU_TREASURE_SHOP_MIX_CROSSLINK_SCOUT_RESULT_2026-06-09.md

v2.60.1