- Introduced a `ProcessingMode` enum to differentiate between basic and premium processing modes.
- Updated `EtlRequest` to include a `processing_mode` field, defaulting to basic.
- Enhanced ETL pipeline services to utilize the selected processing mode for Azure Document Intelligence and LlamaCloud parsing.
- Modified various routes and services to handle processing mode, affecting document upload and indexing tasks.
- Improved error handling and logging to include processing mode details.
- Added tests to validate processing mode functionality and its impact on ETL operations.
Enhanced the NewChatPage to utilize the new GenericHitlApprovalToolUI for handling interrupt results. Updated the ToolFallback component to conditionally render the approval UI based on the result type. Additionally, introduced a new GenericHitlApprovalToolUI component to manage user approvals and parameter editing for tool actions.
Convert MobileEditorPanel, MobileHitlEditPanel, and MobileReportPanel from
static imports to dynamic imports with ssr: false. These panels pull in heavy
dependencies (PlateEditor, MarkdownViewer) that are only needed when users
interact with them on mobile.
Benefits:
- Reduces initial bundle size for chat page
- Mobile panels load only when opened by user
- Faster initial page load and Time to Interactive
- Heavy editor dependencies are code-split into separate chunks
Fixes#1146
Components were calling useSearchParams() at the top level but only
reading the value inside useEffect or callbacks, never in JSX. This
subscribed the entire component tree to every URL query change.
Fix: read from window.location.search directly inside the effect so
no React subscription is created.
Changes:
- new-chat/page.tsx: read commentId inside effect + popstate listener
for SPA back/forward support; removes subscription from 1500+ line tree
- dashboard/page.tsx: read window.location.search at redirect time;
removes searchParams from dep array
- public-chat-footer.tsx: one-shot mount read for action=clone param
- TokenHandler.tsx: one-shot mount read for token + refresh_token params
Implements Vercel React Best Practices Rule: rerender-defer-reads (5.2)
Resolve 8 conflicts:
- Accept upstream deletion of 3 composio_*_connector.py (unified Google connectors)
- Accept our deletion of ElectricProvider.tsx, use-connectors-electric.ts,
use-messages-electric.ts (replaced by Zero equivalents)
- Keep both new deps in package.json (@rocicorp/zero + @slate-serializers/html)
- Regenerate pnpm-lock.yaml
