feat(i18n): localize system UI and move Ctrl+K to overflow

fix(i18n): wire locale behavior and restyle language picker

feat(wave6): voice TTS, share comments, tags+favorites, export+shortcuts

feat(wave5): mind-map, templates, personas marketplace, meeting mode, pinned, notes, link-preview

feat(wave4): 8 features — slash cmds, search, qr-share, /help, image gen, code sandbox, debate, cost tracker

feat: message branching + team workspace (Gist-based collab)

feat: 10+ features — voice/dragdrop/shortcuts, Ollama+RAG citations, workflows+scheduled, share password+TTL, auto-memory, export-md

feat: doc editor + offline mode + stats trio + i18n framework

feat(tools): 3 new tools - Jina Reader URL fetch + YouTube transcript + SQL/CSV analysis

feat(webSearch): multi-provider chain (Brave/Tavily/Serper/Jina) + free fallback

fix(tools): currentTime accepts timezone+locale, fixes Gemini null-args bug

feat(providers): add OpenRouter + Cerebras + Mistral free providers

fix(rate-limits): correct Groq + GitHub Models entries against official docs

fix(rate-limits): Gemma is unlimited on Google free tier (RPM/TPM only)

feat(generateImage): auto-enhance short/non-English prompts via Groq

fix: drop generateImage pre-warm + add loading skeleton

feat: real provider rate-limit data via response headers

fix(generateImage): pre-warm Pollinations cache + better instructions

fix(model-store): replace state ref on each notify so React re-renders

feat(model-picker): shared store + rate-limit badges

fix(mobile): lock body to dvh + min-h-0 main to stop composer scroll-up

feat(cloud-sync): GitHub OAuth + Upstash Redis cross-device sync (LWW)

fix(tools+models): repair UTF-8 mojibake in tool descriptions; trim model list to verified-working set; filter non-chat models from Google live discovery

fix(google): remove unsupported Gemma static IDs; rely on live discovery

feat(google): expand model list w/ Gemma-3 + live discovery /api/google/models

fix(critique): strip <think> reasoning blocks from self-critique output

feat(tools): webSearch fallback chain (Brave -> DDG html -> DDG lite -> Wikipedia)

feat(wave10): A/B compare + diff view + self-critique + z-index hierarchy fix

feat(server): smart context compaction when near limit

feat(composer): image markup with canvas draw tools

feat(ui): density toggle + reading mode

feat(rag): client-side context prefetch from active collection

feat(tools): add runJs, crypto, stock, translate, github, email tools

feat(rag): document collections with mini-RAG via transformers.js

feat(snippets): reusable text macros with variable expansion

feat(themes): custom accent color picker

feat(tools): runPython with Pyodide WASM

feat(message): citations, code enhancements, reactions, chart/mermaid wiring

feat(voice): hands-free voice mode with auto STT/TTS

feat(router): add auto model router based on prompt heuristics

feat(tools): add chartGen + mermaid tools

feat(phase1-wave5): SSE streaming + share-link + BYOK + analytics + branching + a11y + nav-bar fix

feat(phase1-wave4): voice + tool-toggle + prompt-library + error-log + fix webSearch DDG

feat(phase1-wave3): pin/tags/FTS + file upload + personas + memory

feat(phase1-wave2): export chat + token counter + shortcuts + PWA + adaptive typewriter

feat(phase1): theme toggle + slash commands + continue button + 4 new tools

fix(models): drop GitHub Llama-3.1 (no tool-calling via Azure endpoint)

fix(models): purge decommissioned models, only keep verified-working list

feat: rebuild UI with Next.js 15 + shadcn/ui + Vercel AI SDK