-
docs: link the agent trace (Sharing is Caring) + clarify the inference log is separate
nextmarte +1: Claude Opus 4.8 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> -
tests, inference trace, English-only repo, testability refactor
Marcus Ramalho +1: Claude Opus 4.8
- tests/test_logic.py: 7 unit tests (command detection incl. STT errors, language choice, quiet check, path) — pytest green
- tests/smoke_e2e.py: full STT/VLM/TTS pipeline check (PT+EN) — passing
- core/trace.py: opt-in inference trace logger (Sharing-is-Caring) + scripts/publish_trace.py + sample
- app.py: extract pure functions (choose_lang, is_quiet) for testing; wire trace logging
- frontend: all comments translated to English (UI strings stay EN/PT)
- SUBMISSION.md translated to English and updated with status
- gitignore generated traces + internal PT strategy notes (kept local)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> -
Iris: hands-free live mode, money/bill reading, accessible UI, Qwen3-VL-2B
Marcus Ramalho +1: Claude Opus 4.8
- live mode: visual-change watch (parsimonious) + continuous hands-free listening (Web Speech) + baseline scene description; double-tap or voice toggle
- voice commands robust to STT errors (fuzzy: 'modo ao fio' -> live_on)
- reads money/bills/medicine (prompt-tuned) ; EN/PT throughout incl. TTS voices
- accessible custom UI: SVG icons, big labelled buttons, high-contrast/larger-text, focus rings, ARIA, haptics
- language chosen by voice on first run
- swapped VLM to Qwen3-VL-2B (lighter, faster, strong OCR; all models <=4B -> Tiny Titan)
- code/comments/docstrings in English; polished README
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> -
Iris: voice-first assistant for the blind (gr.Server + Qwen2.5-VL + Whisper + Piper)
Marcus Ramalho +1: Claude Opus 4.8
- gr.Server backend: /describe API (image + optional voice -> PT description + speech)
- voice-first frontend: live camera, tap-to-describe / hold-to-ask, high-contrast, EN/PT
- core/ pipeline vendored (STT Whisper, VLM Qwen2.5-VL-3B, TTS Piper pt_BR)
- runs in-Space on ZeroGPU, <=32B total, no third-party model APIs
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>