LettuceAI/app desktop-dev-180-1-786a3d1

Desktop Dev Build #180.1 (786a3d1)

on GitHub

Automated desktop build artifacts for commit 786a3d1c507051d15f131b3dbe6d145684ae6bd3.
Workflow run: https://github.com/LettuceAI/app/actions/runs/26217234057

Changes since previous Desktop build desktop-dev-178-1-af4de66

Compare: desktop-dev-178-1-af4de66...786a3d1

• 786a3d1 feat(sync): unify local model requirements prompts
• 40c9183 fix(sync,onboarding): surface embedding prompt after sync completion
• 7f0bc39 fix(sync,onboarding): reapply handshake and post-sync completion fixes
• 649337e fix(ci): support macOS sha256 generation
• a1dcec7 docs(changelog): add 1.8.1 / 1.5.1 release entry
• ebb7cbc refactor(chat-settings): fetch character by id on reopen
• b731a6b fix(llama.cpp): avoid startup toast for reused models
• e436b81 feat(downloads): show ETA in queue cards
• de281bd fix(memory): fail stalled embedding migrations cleanly
• 25fef58 fix(android): support chat imports from document URIs
• a7cf096 fix(model-browser): suppress GPU-optimal warning for RAM KV cache
• 4dd201c fix(ci): publish checksums for release artifacts
• 9282a8b fix(chat): broadcast updates after toggling message pin
• bffe76a fix: remove unused onboarding model prop
• 27625f4 chore: remove dead code warnings
• 9330b0f feat(onboarding): polish desktop flow contrast, defaults, and step switching
• 6c17b93 docs(changelog): add 1.8.0 and 1.5.0 entry
• 0f67403 fix(chat-appearance): persist overrides and honor background styling
• 0cba8a3 Repair character schema on startup
• 88a2ac1 chore(android): drop forced JNI debug symbols from debug APK
• ab12daf feat(onboarding): add "Sync from another device" flow with embedding check
• a154c17 feat(sync): add Ready handshake and harden passenger apply failures
• e22fd48 fix(hf-browser): honor manual KV cache placement in runnability
• 4104724 perf(ui): adopt NumberInput in EditModelPage
• cf53a07 fix(settings): preserve advanced settings in sync and backups
• b1451c4 feat(ui): add NumberInput primitive and convert clamp-on-keystroke inputs
• 5736c8a feat(settings): add prompt overrides for reply helper
• 5cdbc92 perf(library): persist list cache to localStorage and defer card gradients
• e6de003 perf(library): lazy-load avatar images
• 571c3de feat(dynamic-memory): support custom summary/memory-manager prompts and refresh shared settings UI
• 0e73020 feat(whats-new): add post-update what's-new drawer
• af7e545 fix(import): preserve banner crop in character transfers
• dd197b7 fix(storage): skip invalid characters during list loading
• 57eb1a1 fix(migrations): repair missing character banner crop columns
• 31f6578 feat(tts): add Fish cloud and local audio providers
• bd75cfb feat(settings): redesign system prompts toolbar with multi-select type filter
• 43a336d feat(settings): reorganize prompt entry editor into sections with collapsible conditions
• 68769a4 feat(settings): redesign system prompt entry card with metadata pills
• ef96fa4 refactor(settings): remove unused quick-insert sidebar from prompt template editor
• c67ace1 fix(memories): allow companion memory categories in session edits
• 77bc657 feat(settings): move chat appearance preview to right column
• 38a797c feat(settings): two-pane desktop layout for color customization
• d315a9d refactor(settings): extract field primitives for provider editor
• 6f62576 feat(settings): improve providers page desktop layout
• 52c3d65 feat(providers): add Cerebras AI provider
• 36fe024 fix(sync,backup): persist memory_embeddings as source-of-truth data
• f2f1dda perf(chats): speed up chats and library page load
• b886e2e feat(onboarding): redesign welcome page and unify with onboarding flow
• 37f38d5 feat: make lorebook trigger preview matches open on click
• fe4c1e5 feat(companion): add shared memory across sessions
• 432db89 feat(companion): initialize time awareness from character defaults
• feec88a refactor(soul-editor): slim generate action and make slider percents editable
• b762147 refactor(switch): replace hidden checkbox with button role
• 3884231 feat(companion): scheduled notes editor in soul tab and chat settings
• b1ae90a feat(companion): inject scheduled notes into prompts and add lorebook condition
• ec5355d feat(companion): add scheduled notes storage, commands, and backup
• 338e8ee refactor(chat-io): replace chatpkg zip format with sillytavern jsonl
• 84ffcc9 fear(provider): add Pollinations AI
• b35ec38 refactor(banner-picker): enlarge preview, reposition action chip, and surface remove control
• 598b69c fix(navigation): exit character editor to home instead of looping into settings
• 590368e feat(characters): add banner card type with image picker and cropper
• bfd3e59 refactor(sync): redesign sync UI with neutral palette and progress panel
• f2025fb fix(sync): clear stale host approval state on disconnect
• e2235ef fix(sync): stabilize progress reporting and asset transfer lifecycle
• b55339a chore(deps): pin tanstack virtual version
• f25a914 feat(ci): add Linux package publishing workflows
• a96bf23 fix(groq): use OpenAI-compatible models endpoint
• 3254364 fix(ui): add Groq provider logo
• 8969418 fix(api): strip visible chat metadata from all provider messages
• f9000e9 fix(api): prevent internal request metadata leaking to providers
• f16901d fix(android): revert reqwest to rustls-tls roots bundle
• 60b7fe5 fix(android): allow text tool-call parsing on mobile
• 22d7f48 chore: hoist dynamic imports already imported statically
• 2dc84cf feat(image-gen): attribute usage to source session and character
• 2b449b8 feat(usage): redesign analytics pages with dense table layout
• e304ee0 chore: rustfmt nits
• f3f7ee5 fix(sync): preserve schema fields in backup and sync
• ef3a986 fix(memory): gate v4 upgrade toast on dynamic memory v3
• f442b90 feat(models): add Ollama mode to model browser and redesign installed models
• 32fe540 chore: fix tailwind and cargo warnings
• 4992703 feat(settings): persistent collapsible sidebar with merged Lorebooks and Companions hubs
• 66551cf feat(companion): add per-session time awareness for memory recall and stamping
• a0da527 refactor(chat): align system-send confirm with bottom menu design language
• fe8e88d feat(settings): redesign settings with persistent sidebar and refreshed pages
• 7d82c60 feat(tls): support trusted certs and self-hosted TLS exceptions
• 021febb perf(color-picker): reduce rerenders during gradient editing
• 5f7fa1b fix(avatar-gradient): add source selection and refresh palette state
• 50af1fd fix(avatar): persist cropped base/round avatars and refresh gradient caches
• 9e97372 fix(avatar-gradient): add gradient regeneration and remove brightness lift
• d5d1b3d fix(creation): use proper uuid fallback for character/persona ids
• d3d73c6 refactor(smart-creator): rebuild as agentic tool-calling flow with inline pills, image gallery, and save-and-chat
• 2650699 refactor(smart-creator): remove lorebook flow and make agent controls explicit
• aa8b072 feat(readme): update keywords
• f657e59 feat(llama): add DRY sampler controls and runtime wiring
• e1e9628 feat(lorebook): add agentic lorebook generator with planner, writer, refine, coherence stages
• ced7522 feat(personas): support attaching lorebooks to personas
• 5e0070b feat(debug): show request body per chat attempt and use provider-declared extra-body keys
• f077c3d fix(i18n): resolve raw model download progress keys
• 63825a7 fix(companion): make soul settings page scrollable
• e568237 fix(android): gate unused tooling JSON parsers behind non-mobile cfg
• 4db937e chore: update llama-cpp-rs to 0.1.146
• 13e8135 fix(i18n): add models download progress label
• e2f2d73 fix(models): hide non-llm providers from selector
• 15314c4 fix(session): persist full advanced model overrides
• b63770d fix(embedding): activate downloaded model version on upgrade
• 75f965f fix(request-builder): isolate provider-specific extra body fields