Repository Public
-
Thu, Nov 19, 202624 commits
-
Local route (MacLustr AI): LOCAL badge, picker and Explorer section, local facts, « local · 0 $ » cost wording, local provider row, tests
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Reference screenshots (light + dark), perf summary fix, 2.0 test report and architecture notes
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Integration: backend 1.0 suite on a minted session (journey opt-in), stop test
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Palette keyboard focus, sync refresh for memory / presets / providers, 2.0 docs (release notes, feature matrix, README)
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Screenshots script and reference screenshots, arena hand-over, constellation layout bound, UI smoke test for 2.0 surfaces
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Hosted tests for the 2.0 chat surface, backend 1.0 integration suite, tool-name order
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Merge Arena, Cost Intelligence and Conversation Constellation
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Variable-price routers handled in Core
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Merge contextual panel (context, timeline + replay, Response Lens, branches); variable-price models
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Tests: lighter 300-node layout timing test
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Merge Explorer, Live Model Intelligence and providers settings
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Core tests for backend 1.0 additions; picker sizing; tool-call folding; dev scripts scoped to this checkout
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Arena, Cost Intelligence and Conversation Constellation screens
···
Arena: any-route model selection with a searchable catalogue popover, side-by-side streaming columns with Lumen presence, metrics strip (estimated before / billed or estimated after), prefer, judge, adopt into a conversation, server export, history, window.arenaModelIds hand-over. Usage: budget editing, spend over time, by route, billed vs estimated with the server's note, by model, by conversation (opens it), by kind and project, custom period. Constellation: Canvas graph with off-main deterministic layout, pan / zoom, hover card, selection, keyboard, verbatim basis legend, accessible list. Hosted tests and snapshots on payloads captured on the dev backend. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Context report: real file fields (chars, pages, index status) and file timeline events from the dev backend
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Model Explorer, Live Model Intelligence sheet, compare table and Settings → Fournisseurs
···
- ModelsScreen(window:) searches and filters the whole catalogue (maker, route, capability, category, price ceiling, context floor, free, open-weight, favourites), sorts, and lists dense Equatable rows with favourite / dock / compare toggles - Detail sheet from GET /api/models/detail: facts, published prices, accepted parameters, figures measured on the account, live endpoints, third-party benchmarks, twins, local benchmark — each block labelled with its source - Compare tray (2-4 models) with a side-by-side table and a hand-off to the Arena - Providers pane: state, health, test, set / rotate / revoke of manageable keys; a key is sent once and never kept, shown or logged - Tests: provider and model-detail decoding on real payloads, key flows on MockTransport, ModelQuery filters / sorts / search, hosted view-model tests and snapshot renders Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Contextual panel: Context Inspector, Timeline + Replay, Response Lens, Branches
···
ContextPanel(chat:window:) with a four-tab header and cross-faded content; each tab has its own observable model driving the real endpoints. Hosted tests on a scripted server and light + dark snapshots of the four tabs. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Command Center, menu commands, assistant settings, prompt variables, native artefact previews, Dock-icon drop, 2.0.0
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Visual System V2 across the chat: floating composer with commands, model picker and dock, transcript with source cards and follow-ups, sync coordinator, advanced search
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Core: contextual panel DTOs and endpoints (context, timeline, tree, lens), lens run accumulator, replay cursor
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Core: Arena events + accumulator, Cost Intelligence fields, Constellation models, force layout
···
Typed ArenaEvent / ArenaAccumulator (pure, tested on a run captured on the dev backend), stored comparisons, judge, adopt and export endpoints; UsageReport gains byRoute, byConversation, byProject, monthly, costBasis and range (tolerant); Constellation DTO + endpoint; deterministic seedable ForceLayout. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Next-gen foundation: catalogue model fields, model query, providers, sync feed, Visual System V2 tokens, logo assets
···
Core: tolerant decoding of route / creator / logo / categories / catalog, OpenRouter catalogue (routes=all), ModelQuery + search index, providers and key management, model detail, route suggestion, sync feed reducer and client, web search, follow-ups, citations, advanced search filters. App: design tokens as asset catalog colours, floating material, ModelMark / ModelChip / RouteBadge / Lumen line, bundled maker logos generated by Scripts/sync-logos.py, CatalogStore over the whole catalogue with favourites, recents and dock. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
docs: link sibling iOS / core repositories in the family table
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
docs: legendary README — icon, logo, live spbgit badges, author
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
-
Tue, Nov 17, 20265 commits
-
notarize.sh: checksum sidecar with the bare file name
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
SP Chat 1.0.0: documentation, test report, release notes, final provisioning profile
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Workspace screens, app icon (SVG), hosted tests, model picker sizing
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Native app shell: environment, session, sidebar, chat, composer, Markdown/LaTeX rendering, model picker, settings, commands, quick chat
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-
Core: models, networking, SSE engine, passkey auth, SQLite cache, Markdown/LaTeX model, tests (72 unit + backend integration)
···
Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
-