- Add GLM-4.7-Flash as default Ollama model (replaces Mistral)
- Fix WRITE_TIMEOUT (15s→120s) and HTMX timeout (5s→120s) for local LLM
- Auto-warmup model on startup in development mode
- Add /api/chat/status endpoint for model readiness polling
- Show "Initializing AI model..." indicator while model loads
- Add user avatar (mdi:account) on chat messages
- Inject company/project/course sprite icons inline in chat responses
- Replace cramped header icons with 4 icon buttons + tooltips
(Compact, Side panel, Floating, Full screen)
- Add floating/draggable chat mode with smooth drag support
- Chip questions show user bubble instantly and clear input
- Help modal prefills input instead of auto-sending
- User bubble rendered client-side for immediate feedback
Dual-provider architecture:
- Both Gemini and Ollama initialize at startup (if configured)
- Primary (Gemini) tried first for every request
- On any error (429, 503, timeout), automatically falls back to Ollama
- No manual switching needed — completely transparent to the user
- Log shows: "Primary failed (gemini: ...), falling back to ollama: ..."
Warmup:
- POST /api/chat/warmup called silently when chat panel opens
- Pre-loads Ollama model in background (10-15s) while user reads welcome
- By the time user types, model is ready for instant response
- Warms up fallback provider specifically (Gemini doesn't need it)
Timeout:
- Agent context increased to 60s (Ollama first response can be slow)
- Each request creates a fresh session (stateless for fallback compat)
Ollama adapter (internal/chat/ollama.go):
- Implements model.LLM interface for ADK Go
- Talks to Ollama's OpenAI-compatible API (/v1/chat/completions)
- Full tool/function calling support (tested with Mistral Small 3.2)
- Converts ADK types to OpenAI format (messages, tools, tool_calls)
- Configurable via OLLAMA_HOST and OLLAMA_MODEL env vars
Multi-provider handler:
- MODEL_PROVIDER env: "gemini" (default) or "ollama"
- Gemini: requires GOOGLE_API_KEY (pay-as-you-go recommended)
- Ollama: connects to local or Tailscale-remote instance
Rate limiter:
- 30 requests/hour per IP on /api/chat endpoint
- Uses existing middleware.NewRateLimiter pattern
Tested: Ollama + Mistral Small 3.2 on M4 Pro 64GB — correct answers
Visitors can ask questions about the CV via a floating chat panel.
The agent uses Gemini to answer questions about experience, projects,
skills, and education by querying the cached CV JSON data.
- internal/chat/agent.go: LLM agent with query_cv tool that searches
CV data by section (experience, projects, skills, etc.) with keyword filtering
- internal/chat/handler.go: POST /api/chat endpoint with session management,
graceful degradation when GOOGLE_API_KEY is not set
- chat-widget.html: HTMX-powered floating chat panel with Hyperscript toggle
- _chat.css: Responsive chat UI with dark theme support
- Wired into existing architecture via dependency injection (CVHandler,
routes, main.go) — zero breaking changes, all existing tests pass
- Merge lang package into constants (add IsValidLang, ValidateLang, AllLangs)
- Rename internal/services to internal/email for consistency with pdf package
- Rename types to avoid redundancy: EmailService→Service, EmailConfig→Config
- Update all imports and references across codebase
- Delete internal/lang directory (functions moved to constants)
- Add aria-labels to menu action buttons (PDF, Print, Contact)
- Add aria-labelledby to toggle checkboxes (desktop + mobile)
- Add -webkit-user-select prefix for Safari compatibility
- Add DynamicCacheControl middleware for HTML pages
- Add accessibility test suite (60-accessibility.test.mjs)
- Add comprehensive accessibility documentation (21-ACCESSIBILITY.md)
- Update Modern Web Techniques doc to mark audit complete
Implement a command palette accessible via CMD+K/Ctrl+K using the ninja-keys
web component. Features include:
- New /api/cmd-k endpoint serving dynamic CV entries (experiences, projects, courses)
- Language-aware responses with 1-hour cache headers
- Scroll-to-section functionality for quick navigation
- Enhanced keyboard shortcuts modal with CMD+K documentation
- Comprehensive test coverage for API and UI interactions
Also includes cleanup of deprecated debug test files and various UI polish
improvements to contact form, themes, and action bar components.
Plain text endpoint:
- Add /text route for plain text CV (for curl/AI crawlers)
- Use k3a/html2text library for HTML-to-text conversion
- Add Plain Text button to hamburger menu with UI translations
Contact form feature:
- Add ContactHandler with proper email service integration
- Add CSRF protection middleware
- Add rate limiting (5 submissions/hour per IP)
- Add honeypot and timing-based bot protection
- Add input validation with detailed error messages
- Add security logging middleware
- Add browser-only middleware for API protection
Code quality:
- Fix all golangci-lint errcheck warnings for w.Write calls
- Remove duplicate getClientIP functions
- Wire up ContactHandler in routes.Setup
Complete middleware integration with comprehensive testing:
1. Middleware Integration
- Added PreferencesMiddleware to middleware chain in routes
- Order: Recovery → Logger → SecurityHeaders → Preferences → Mux
- Reads all preference cookies once per request
- Stores in context for handlers to access
2. Handler Updates
- cv_pages.go: Home handler uses middleware.GetPreferences()
- cv_htmx.go: All toggle handlers use middleware preferences
- Eliminated manual cookie reading in handlers
- Migration logic handled entirely by middleware
3. Comprehensive Middleware Tests
- Created preferences_test.go with 10+ test functions
- Tests: default values, migrations, cookie setting, context access
- Verified: extended→long, true→show, false→hide migrations
- All tests passing
Benefits:
- Performance: Cookies read once per request (not multiple times)
- Consistency: All handlers get same preference values
- Maintainability: Migration logic centralized in middleware
- Testability: Easy to mock preferences via context
Testing:
- All unit tests pass (handlers + middleware)
- Build succeeds
- No breaking changes
## Shortcut URLs
- New routes: /cv-jamr-{year}-{lang}.pdf (e.g., /cv-jamr-2025-en.pdf)
- Year validation: Only current year accepted, returns 404 for past/future
- Auto-redirects (301) to: /export/pdf?lang={lang}&length=short&icons=show&version=with_skills
- Both languages supported: en and es
## PDF Modal Updates
- Replaced "Current View" option with "Default CV (Recommended)"
- Visual highlighting: purple gradient badge, star emoji ⭐, bold text
- Uses shortcut URL with dynamic year detection
- Clear recommendation for users (5 pages, short with skills)
## Technical Details
- Handler: DefaultCVShortcut() in internal/handlers/cv.go
- Pattern check in Home() handler for proper routing
- Helper function: window.openPdfModal() for references section
- Documentation: PDF-SHORTCUT-IMPLEMENTATION.md
Benefits:
- Memorable, shareable URLs (juan.andres.morenorub.io/cv-jamr-2025-en.pdf)
- Auto-updates yearly without code changes
- Clear user guidance for recommended CV format
- Removed over-engineered cache system for static CV data that only changes on deployment
- Extracted all route configuration to internal/routes/routes.go for better organization
- Implemented rate limiting and cache control middleware for PDF endpoint protection
juanatsap