LettuceAI/app desktop-dev-169-1-51dc919

Desktop Dev Build #169.1 (51dc919)

on GitHub

Automated desktop build artifacts for commit 51dc919a82ece12701fe724dfe8c69800790ddb5.
Workflow run: https://github.com/LettuceAI/app/actions/runs/24670373614

Changes since previous Desktop build desktop-dev-168-1-54479e6

Compare: desktop-dev-168-1-54479e6...51dc919

• 51dc919 feat(lorebook): mobile trigger preview with tabbed layout
• 1d0faac feat(lorebook): redesign metadata menu with library image picker
• c2d5716 feat: expand attached lorebook editor actions
• 49c4a5a feat(lorebook): redesign list view with active/available sections
• aedb085 refactor(ui): simplify lorebook avatar fallback
• cc46404 feat: add seeded lorebook benchmark test
• 9b33834 refactor(lorebook): swap inline preview for route button in editors
• d2e0317 feat(lorebook): add dedicated trigger preview page
• 7a8ca6d feat(tokens): add o200k_base tokenizer command
• c642e7c fix: prevent lorebook search header overlap
• be57291 fix: use selected variants in scene prompts
• eab0d16 feat: add scene prompt writer template
• 91bcf3d fix: show help me reply reasoning status
• b89fe4d feat(chat): add guided regeneration
• 0b44dc4 feat(lorebooks): centralize active lorebook overrides
• 385fd32 feat(providers): add opt-in reasoning sync for custom endpoints (#50)
• 2a99fd0 fix(chat): keep all assistant regenerate variants
• 5f85383 feat(chat): add attachment edit sheets and compact scene prompt previews
• 9ca1ea0 fix(sync): reset stale payload state after schema drift
• 7f81145 feat(chat): allow removing message attachments during edit
• 03cd33b Merge pull request #48 from Husky110/feat/message-debug-view-improvements
• 2e21bc5 fix(dynamic-memory): limit llama sampler overrides to llama.cpp
• b3141ad fix(group-chat): load assigned character prompt templates from live columns
• 4e0616a Merge branch 'LettuceAI:main' into feat/message-debug-view-improvements
• cedc7be feat(debug): improve message debug view with collapsible sections and design tokens
• 405f25e fix(llama-cpp): disable gpu-only controls on cpu backends
• ffcbf37 fix(models): use ram-only runability scoring on cpu backends
• 05f6344 fix(updater): detect 0.1.x windows builds as dev channel
• 27d4042 fix(chat): persist edited session scene overrides
• d588b34 fix(branching): preserve memory state when branching group chats to direct
• 7c5734b fix(branching): rebuild visible memory state from kept dynamic events