Skip to content

Feat/v2 spotlight#2

Merged
createpjf merged 5 commits intomainfrom
feat/v2-spotlight
Mar 4, 2026
Merged

Feat/v2 spotlight#2
createpjf merged 5 commits intomainfrom
feat/v2-spotlight

Conversation

@createpjf
Copy link
Owner

  1. Add Local Model Support(Ollama and LM Studio)
  2. Add Web Search Function(Brave, needs API Key)
  3. Add Playground
  4. UI/UX Improvement

createpjf and others added 5 commits March 4, 2026 11:59
- Add Spotlight window (⌘⇧X): floating quick-ask with streaming markdown
- Add Chat window: multi-turn conversations with sidebar, compare mode
- Add clipboard shortcuts: ⌘⇧T translate, ⌘⇧S summarize, ⌘⇧E explain
- Add My Usage tab: today/month stats, 7-day trend, model breakdown
- Inject routebox.meta in SSE streams for per-response cost tracking
- Add conversations/messages/spotlight_history SQLite tables + CRUD
- Add 13 new API endpoints (usage, conversations, spotlight)
- Hash-based multi-window routing with React.lazy code splitting
- Shared components: MarkdownRenderer, ModelSwitcher, CostBar, useStreamChat

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds MessageSquare (Chat) and Sparkles (Spotlight) icons to the
HeroSection title bar so users can open these windows from the panel.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Auto-discover local LLM servers at localhost:11434 and localhost:1234
- Dynamic model listing via /v1/models probe (30s polling)
- $0 pricing for all local inference, 120s timeout for cold starts
- Skip auth headers for local providers
- 3 new API endpoints for local provider management
- Frontend UI with status dots, model count, URL editing, refresh
- cost_first routing naturally prefers local ($0)
- Provider colors: Ollama (#0077B6), LM Studio (#E63946)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
New features:
- Brave Search integration: real-time web search in Chat and Spotlight
- Full Chat window with conversation history and sidebar
- Thinking model support (DeepSeek-R1, Qwen3) with collapsible blocks
- ModelSwitcher now only shows providers with active API keys
- Auth token auto-loaded across all windows via Keychain

Bug fixes:
- Fix budget sync bug: setting key mismatch (monthly_budget → budgetMonthly)
- Fix auth token not propagating to Chat/Spotlight windows

Infrastructure:
- Updated README with v2.3.0 features and Brave Search docs
- Added search status/key management API endpoints
- Search results injected as system context before LLM forwarding

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@createpjf createpjf merged commit bd4867e into main Mar 4, 2026
2 of 3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant