fix: sanitize ICD-10 codes + UI/chat history fixes

refactor: system/user prompt split + robust JSON extraction

fix: localize condition names + add edge-tts dependency

feat: TTS listen button + follow-up Q&A chat

refactor: split format_agent into 2 separate LLM calls (patient + SOAP)

fix: LLM ignores ===SOAP=== delimiter — stronger prompt + fallback parser

fix: patient message too short — enforce 5-7 sentence minimum in FORMAT_AGENT_SYSTEM

fix: SOAP Subjective always English, tab labels selector robustness

fix: 7 bugs — PART labels, hallucination, lang switch, tab i18n, SVG click, img-mode radio, actions label

fix: format agent prompt, SVG click JS, tab label i18n

feat: 3-step agentic pipeline — A/B comparison, red-flag triage, SOAP note

feat: Anatomical Region Selector — body map + dropdown inject region into prompt

feat: AMD Live Inference Dashboard — show latency, throughput, token count after each analysis

fix: enforce target-language output for diagnosis/actions, keep severity in English for stable parsing

fix: move language selector global + fix diagnosis prompt echo bug

feat: multilingual support (ZH/ES/FR/JA) + remove all mock mode

fix: add detailed connection logs to diagnose AMD Cloud failures

feat: add AMD Cloud connection status bar, replace MOCK badge

refactor: switch to AMD Dev Cloud vLLM backend with Qwen2.5-VL

Production update: full Gradio app with AMD branding, mock+real mode, bilingual support