The deferred tool response approach caused Gemini to timeout waiting.
Now request_contact responds immediately (telling the agent to wait),
and the confirm button sends a text message through the live WebSocket
to trigger complete_brief.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- request_contact tool shows on-screen card for name/email verification
- Deferred tool responses let the UI wait for user confirmation
- WebSocket close preserves transcript and enables reconnection
- Reconnect seeds new Gemini session with prior conversation context
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Agent now asks probing follow-up questions (pain points, current tools, vision, decision context)
- complete_brief tool includes conversationSummary field capturing ALL discussion details
- Summary merges into scope context for richer, more targeted briefs
- Updated both EN and FR system prompts
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Visual feedback while the complete_brief API call runs (~15s).
Badge appears below the orb with a spinner animation.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- briefSubmittedRef prevents multiple complete_brief calls
- Added logging throughout completion flow for debugging
- Reset ref on new conversation start
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Streaming transcription words now merge into one bubble per agent/user turn
instead of creating separate entries for each chunk.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Introduction prompt uses realtimeInput.text format (not clientContent)
- Agent now successfully introduces itself and sends audio responses
- Playback AudioContext created in user gesture for mobile
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Setup uses "generationConfig" field (not "config") inside "setup" object
- Handle Blob data from WebSocket (Gemini sends binary, not text)
- Voice agent now successfully connects and receives setupComplete
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Rate limit reduced from 30s to 5s for testing
- WebSocket setup times out after 10s instead of hanging forever
- Added connection logging
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Site analysis: cheerio HTML parsing, inline tech stack detection (~20 CMS/framework/analytics signatures), Google PageSpeed API integration
- Gemini Live voice agent: WebSocket-based real-time voice mode with live transcript, selection chips, and mid-conversation website analysis
- Type/Talk mode toggle with silent capability detection
- Stepped progress animation during brief generation (4 animated steps)
- URL + thoughts fields in Step 2, phone + contact preference in Step 3
- AI prompt improvements: dedicated website analysis section, 30-min call, concrete benefits, industry depth
- Email redesign: branded templates with logo, proper markdown rendering for both client and admin
- French locale support for AI-generated briefs
- Smaller checkmark, compact booking CTA, expanded brief area
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
