- Show prominent app URL notification when task completes
- Play completion sound using Web Audio API
- Extract app_url from task completion details
- Add slideInRight/slideOutRight animations
- Make toast clickable to open app URL
- Refresh task details on completion
- Add prominent input box at top of tasks page
- Gradient header with clean design
- HTMX integration for /api/autotask/create
- Show progress indicator during processing
- Display result with app link when done
- Enter key triggers submit
- Auto-refresh task list after creation
- Delete alpinejs.min.js from ui/suite/js/vendor/
- Delete alpinejs.min.js from ui/minimal/js/vendor/
- Remove script tag from default.gbui
- Remove [x-cloak] CSS rule from global.css
- Update PROMPT.md to remove Alpine.js documentation
- Add organizations.html with org selector and bot hierarchy tree
- Add bot configuration form with enabled tabs
- Update PROMPT.md with weekly maintenance checklist
- Added Repositories tab with GitHub/GitLab/Bitbucket support
- Added Apps tab for previously created HTMX apps
- @mention autocomplete in chat for repos (@botserver) and apps (@myapp)
- Task context storage for autonomous task execution
- CSS for repo and app cards with connection status
- Mention suggestions dropdown with keyboard navigation
- autotask.html: Auto Task interface with intent input, plan preview, decisions
- autotask.css: Premium styling for Auto Task system
- autotask.js: Client-side logic, WebSocket handling, approval workflow
- Update chat.css to use consistent CSS variable names from app.css
- Changed --primary-color to --accent-color
- Changed --surface-color to --secondary-bg
- Changed --text-color to --text-primary
- Changed --background-color to --primary-bg
- Changed --primary-hover to --accent-hover
- Ensures proper theming across all 6 themes
- Real LLM API calls for insights, sentiment, smart replies
- Polish button (✨) to improve messages before sending
- Tips display when customer messages arrive
- Sentiment analysis with escalation warnings
- Smart reply suggestions from LLM
- Conversation history tracking for context
New rules added:
- Two LLM work modes: Execution (fazer) vs Review (conferir)
- HTMX-first frontend: minimize JS at maximum
- Local assets only: no CDN references
- Server returns HTML fragments, not JSON
- No dead code: implement real code, never use _ prefix
- cargo audit must pass with 0 warnings
- Documentation rules: Rust code only in architecture docs
- Added HTMX patterns and examples
- AppState now uses BotServerClient directly
- BOTSERVER_URL env var support for configuration
- index() handler properly integrated into router
- Removed unused web module (DTOs were never used)
- Removed all #[allow(dead_code)] attributes
- Zero warnings, cargo audit clean
