fix: OpenAI reasoning/thinking blocks not showing

- Use Responses API instead of Chat Completions API for OpenAI
  (.chat() -> default call) to support reasoning events
- Add o4 to reasoning model detection
- Change default reasoningSummary from 'detailed' to 'auto'
  (not all models support 'detailed')
- Update types to match AI SDK: 'auto' | 'detailed'

app/api/chat/route.ts

@@ -14,11 +14,6 @@ import path from "path"
 import { z } from "zod"
 import { getAIModel, supportsPromptCaching } from "@/lib/ai-providers"
 import { findCachedResponse } from "@/lib/cached-responses"
-import {
-    checkAndIncrementRequest,
-    isQuotaEnabled,
-    recordTokenUsage,
-} from "@/lib/dynamo-quota-manager"
 import {
     getTelemetryConfig,
     setTraceInput,
@@ -167,13 +162,9 @@ async function handleChatRequest(req: Request): Promise<Response> {
 
     const { messages, xml, previousXml, sessionId } = await req.json()
 
-    // Get user IP for Langfuse tracking (hashed for privacy)
+    // Get user IP for Langfuse tracking
     const forwardedFor = req.headers.get("x-forwarded-for")
-    const rawIp = forwardedFor?.split(",")[0]?.trim() || "anonymous"
+    const userId = forwardedFor?.split(",")[0]?.trim() || "anonymous"
-    const userId =
-        rawIp === "anonymous"
-            ? rawIp
-            : `user-${Buffer.from(rawIp).toString("base64url").slice(0, 8)}`
 
     // Validate sessionId for Langfuse (must be string, max 200 chars)
     const validSessionId =
@@ -182,12 +173,9 @@ async function handleChatRequest(req: Request): Promise<Response> {
             : undefined
 
     // Extract user input text for Langfuse trace
-    // Find the last USER message, not just the last message (which could be assistant in multi-step tool flows)
+    const lastMessage = messages[messages.length - 1]
-    const lastUserMessage = [...messages]
-        .reverse()
-        .find((m: any) => m.role === "user")
     const userInputText =
-        lastUserMessage?.parts?.find((p: any) => p.type === "text")?.text || ""
+        lastMessage?.parts?.find((p: any) => p.type === "text")?.text || ""
 
     // Update Langfuse trace with input, session, and user
     setTraceInput({
@@ -196,33 +184,6 @@ async function handleChatRequest(req: Request): Promise<Response> {
         userId: userId,
     })
 
-    // === SERVER-SIDE QUOTA CHECK START ===
-    // Quota is opt-in: only enabled when DYNAMODB_QUOTA_TABLE env var is set
-    const hasOwnApiKey = !!(
-        req.headers.get("x-ai-provider") && req.headers.get("x-ai-api-key")
-    )
-
-    // Skip quota check if: quota disabled, user has own API key, or is anonymous
-    if (isQuotaEnabled() && !hasOwnApiKey && userId !== "anonymous") {
-        const quotaCheck = await checkAndIncrementRequest(userId, {
-            requests: Number(process.env.DAILY_REQUEST_LIMIT) || 10,
-            tokens: Number(process.env.DAILY_TOKEN_LIMIT) || 200000,
-            tpm: Number(process.env.TPM_LIMIT) || 20000,
-        })
-        if (!quotaCheck.allowed) {
-            return Response.json(
-                {
-                    error: quotaCheck.error,
-                    type: quotaCheck.type,
-                    used: quotaCheck.used,
-                    limit: quotaCheck.limit,
-                },
-                { status: 429 },
-            )
-        }
-    }
-    // === SERVER-SIDE QUOTA CHECK END ===
-
     // === FILE VALIDATION START ===
     const fileValidation = validateFileParts(messages)
     if (!fileValidation.valid) {
@@ -276,10 +237,9 @@ async function handleChatRequest(req: Request): Promise<Response> {
     // Get the appropriate system prompt based on model (extended for Opus/Haiku 4.5)
     const systemMessage = getSystemPrompt(modelId, minimalStyle)
 
-    // Extract file parts (images) from the last user message
+    // Extract file parts (images) from the last message
     const fileParts =
-        lastUserMessage?.parts?.filter((part: any) => part.type === "file") ||
+        lastMessage.parts?.filter((part: any) => part.type === "file") || []
-        []
 
     // User input only - XML is now in a separate cached system message
     const formattedUserInput = `User input:
@@ -288,7 +248,7 @@ ${userInputText}
 """`
 
     // Convert UIMessages to ModelMessages and add system message
-    const modelMessages = await convertToModelMessages(messages)
+    const modelMessages = convertToModelMessages(messages)
 
     // DEBUG: Log incoming messages structure
     console.log("[route.ts] Incoming messages count:", messages.length)
@@ -542,26 +502,12 @@ ${userInputText}
                 userId,
             }),
         }),
-        onFinish: ({ text, totalUsage }) => {
+        onFinish: ({ text, usage }) => {
-            // AI SDK 6 telemetry auto-reports token usage on its spans
+            // Pass usage to Langfuse (Bedrock streaming doesn't auto-report tokens to telemetry)
-            setTraceOutput(text)
+            setTraceOutput(text, {
-
+                promptTokens: usage?.inputTokens,
-            // Record token usage for server-side quota tracking (if enabled)
+                completionTokens: usage?.outputTokens,
-            // Use totalUsage (cumulative across all steps) instead of usage (final step only)
+            })
-            // Include all 4 token types: input, output, cache read, cache write
-            if (
-                isQuotaEnabled() &&
-                !hasOwnApiKey &&
-                userId !== "anonymous" &&
-                totalUsage
-            ) {
-                const totalTokens =
-                    (totalUsage.inputTokens || 0) +
-                    (totalUsage.outputTokens || 0) +
-                    (totalUsage.cachedInputTokens || 0) +
-                    (totalUsage.inputTokenDetails?.cacheWriteTokens || 0)
-                recordTokenUsage(userId, totalTokens)
-            }
         },
         tools: {
             // Client-side tool that will be executed on the client
@@ -731,9 +677,19 @@ Call this tool to get shape names and usage syntax for a specific library.`,
         messageMetadata: ({ part }) => {
             if (part.type === "finish") {
                 const usage = (part as any).totalUsage
-                // AI SDK 6 provides totalTokens directly
+                if (!usage) {
+                    console.warn(
+                        "[messageMetadata] No usage data in finish part",
+                    )
+                    return undefined
+                }
+                // Total input = non-cached + cached (these are separate counts)
+                // Note: cacheWriteInputTokens is not available on finish part
+                const totalInputTokens =
+                    (usage.inputTokens ?? 0) + (usage.cachedInputTokens ?? 0)
                 return {
-                    totalTokens: usage?.totalTokens ?? 0,
+                    inputTokens: totalInputTokens,
+                    outputTokens: usage.outputTokens ?? 0,
                     finishReason: (part as any).finishReason,
                 }
             }

app/api/log-feedback/route.ts

@@ -27,18 +27,9 @@ export async function POST(req: Request) {
 
     const { messageId, feedback, sessionId } = data
 
-    // Skip logging if no sessionId - prevents attaching to wrong user's trace
+    // Get user IP for tracking
-    if (!sessionId) {
-        return Response.json({ success: true, logged: false })
-    }
-
-    // Get user IP for tracking (hashed for privacy)
     const forwardedFor = req.headers.get("x-forwarded-for")
-    const rawIp = forwardedFor?.split(",")[0]?.trim() || "anonymous"
+    const userId = forwardedFor?.split(",")[0]?.trim() || "anonymous"
-    const userId =
-        rawIp === "anonymous"
-            ? rawIp
-            : `user-${Buffer.from(rawIp).toString("base64url").slice(0, 8)}`
 
     try {
         // Find the most recent chat trace for this session to attach the score to

app/api/log-save/route.ts

@@ -27,11 +27,6 @@ export async function POST(req: Request) {
 
     const { filename, format, sessionId } = data
 
-    // Skip logging if no sessionId - prevents attaching to wrong user's trace
-    if (!sessionId) {
-        return Response.json({ success: true, logged: false })
-    }
-
     try {
         const timestamp = new Date().toISOString()
 

app/globals.css

@@ -144,68 +144,6 @@
     --sidebar-ring: oklch(0.7 0.16 265);
 }
 
-/* ============================================
-   REFINED MINIMAL DESIGN SYSTEM
-   ============================================ */
-
-:root {
-    /* Surface layers for depth */
-    --surface-0: oklch(1 0 0);
-    --surface-1: oklch(0.985 0.002 240);
-    --surface-2: oklch(0.97 0.004 240);
-    --surface-elevated: oklch(1 0 0);
-
-    /* Subtle borders */
-    --border-subtle: oklch(0.94 0.008 260);
-    --border-default: oklch(0.91 0.012 260);
-
-    /* Interactive states */
-    --interactive-hover: oklch(0.96 0.015 260);
-    --interactive-active: oklch(0.93 0.02 265);
-
-    /* Success state */
-    --success: oklch(0.65 0.18 145);
-    --success-muted: oklch(0.95 0.03 145);
-
-    /* Animation timing */
-    --duration-fast: 120ms;
-    --duration-normal: 200ms;
-    --duration-slow: 300ms;
-    --ease-out: cubic-bezier(0.16, 1, 0.3, 1);
-    --ease-in-out: cubic-bezier(0.4, 0, 0.2, 1);
-    --ease-spring: cubic-bezier(0.34, 1.56, 0.64, 1);
-}
-
-.dark {
-    --surface-0: oklch(0.15 0.015 260);
-    --surface-1: oklch(0.18 0.015 260);
-    --surface-2: oklch(0.22 0.015 260);
-    --surface-elevated: oklch(0.25 0.015 260);
-
-    --border-subtle: oklch(0.25 0.012 260);
-    --border-default: oklch(0.3 0.015 260);
-
-    --interactive-hover: oklch(0.25 0.02 265);
-    --interactive-active: oklch(0.3 0.025 270);
-
-    --success: oklch(0.7 0.16 145);
-    --success-muted: oklch(0.25 0.04 145);
-}
-
-/* Expose surface colors to Tailwind */
-@theme inline {
-    --color-surface-0: var(--surface-0);
-    --color-surface-1: var(--surface-1);
-    --color-surface-2: var(--surface-2);
-    --color-surface-elevated: var(--surface-elevated);
-    --color-border-subtle: var(--border-subtle);
-    --color-border-default: var(--border-default);
-    --color-interactive-hover: var(--interactive-hover);
-    --color-interactive-active: var(--interactive-active);
-    --color-success: var(--success);
-    --color-success-muted: var(--success-muted);
-}
-
 @layer base {
     * {
         @apply border-border outline-ring/50;
@@ -319,83 +257,3 @@
     -webkit-text-fill-color: transparent;
     background-clip: text;
 }
-
-/* ============================================
-   REFINED DIALOG STYLES
-   ============================================ */
-
-/* Refined dialog shadow - multi-layer soft shadow */
-.shadow-dialog {
-    box-shadow:
-        0 0 0 1px oklch(0 0 0 / 0.03),
-        0 2px 4px oklch(0 0 0 / 0.02),
-        0 12px 24px oklch(0 0 0 / 0.06),
-        0 24px 48px oklch(0 0 0 / 0.04);
-}
-
-.dark .shadow-dialog {
-    box-shadow:
-        0 0 0 1px oklch(1 0 0 / 0.05),
-        0 2px 4px oklch(0 0 0 / 0.2),
-        0 12px 24px oklch(0 0 0 / 0.3),
-        0 24px 48px oklch(0 0 0 / 0.2);
-}
-
-/* Dialog animations */
-@keyframes dialog-in {
-    from {
-        opacity: 0;
-        transform: translate(-50%, -48%) scale(0.96);
-    }
-    to {
-        opacity: 1;
-        transform: translate(-50%, -50%) scale(1);
-    }
-}
-
-@keyframes dialog-out {
-    from {
-        opacity: 1;
-        transform: translate(-50%, -50%) scale(1);
-    }
-    to {
-        opacity: 0;
-        transform: translate(-50%, -48%) scale(0.96);
-    }
-}
-
-.animate-dialog-in {
-    animation: dialog-in var(--duration-normal) var(--ease-out) forwards;
-}
-
-.animate-dialog-out {
-    animation: dialog-out 150ms var(--ease-out) forwards;
-}
-
-/* Check pop animation for validation success */
-@keyframes check-pop {
-    0% {
-        transform: scale(0.8);
-        opacity: 0;
-    }
-    50% {
-        transform: scale(1.1);
-    }
-    100% {
-        transform: scale(1);
-        opacity: 1;
-    }
-}
-
-.animate-check-pop {
-    animation: check-pop 0.25s var(--ease-spring) forwards;
-}
-
-/* Reduced motion support */
-@media (prefers-reduced-motion: reduce) {
-    .animate-dialog-in,
-    .animate-dialog-out,
-    .animate-check-pop {
-        animation: none;
-    }
-}

components/chat-message-display.tsx

@@ -31,7 +31,6 @@ import { getApiEndpoint } from "@/lib/base-path"
 import {
     applyDiagramOperations,
     convertToLegalXml,
-    extractCompleteMxCells,
     isMxCellXmlComplete,
     replaceNodes,
     validateAndFixXml,
@@ -316,28 +315,12 @@ export function ChatMessageDisplay({
 
     const handleDisplayChart = useCallback(
         (xml: string, showToast = false) => {
-            let currentXml = xml || ""
+            const currentXml = xml || ""
-            const startTime = performance.now()
-
-            // During streaming (showToast=false), extract only complete mxCell elements
-            // This allows progressive rendering even with partial/incomplete trailing XML
-            if (!showToast) {
-                const completeCells = extractCompleteMxCells(currentXml)
-                if (!completeCells) {
-                    return
-                }
-                currentXml = completeCells
-            }
-
             const convertedXml = convertToLegalXml(currentXml)
             if (convertedXml !== previousXML.current) {
                 // Parse and validate XML BEFORE calling replaceNodes
                 const parser = new DOMParser()
-                // Wrap in root element for parsing multiple mxCell elements
+                const testDoc = parser.parseFromString(convertedXml, "text/xml")
-                const testDoc = parser.parseFromString(
-                    `<root>${convertedXml}</root>`,
-                    "text/xml",
-                )
                 const parseError = testDoc.querySelector("parsererror")
 
                 if (parseError) {
@@ -364,22 +347,7 @@ export function ChatMessageDisplay({
                         `<mxfile><diagram name="Page-1" id="page-1"><mxGraphModel><root><mxCell id="0"/><mxCell id="1" parent="0"/></root></mxGraphModel></diagram></mxfile>`
                     const replacedXML = replaceNodes(baseXML, convertedXml)
 
-                    const xmlProcessTime = performance.now() - startTime
+                    // Validate and auto-fix the XML
-
-                    // During streaming (showToast=false), skip heavy validation for lower latency
-                    // The quick DOM parse check above catches malformed XML
-                    // Full validation runs on final output (showToast=true)
-                    if (!showToast) {
-                        previousXML.current = convertedXml
-                        const loadStartTime = performance.now()
-                        onDisplayChart(replacedXML, true)
-                        console.log(
-                            `[Streaming] XML processing: ${xmlProcessTime.toFixed(1)}ms, drawio load: ${(performance.now() - loadStartTime).toFixed(1)}ms`,
-                        )
-                        return
-                    }
-
-                    // Final output: run full validation and auto-fix
                     const validation = validateAndFixXml(replacedXML)
                     if (validation.valid) {
                         previousXML.current = convertedXml
@@ -392,20 +360,19 @@ export function ChatMessageDisplay({
                             )
                         }
                         // Skip validation in loadDiagram since we already validated above
-                        const loadStartTime = performance.now()
                         onDisplayChart(xmlToLoad, true)
-                        console.log(
-                            `[Final] XML processing: ${xmlProcessTime.toFixed(1)}ms, validation+load: ${(performance.now() - loadStartTime).toFixed(1)}ms`,
-                        )
                     } else {
                         console.error(
                             "[ChatMessageDisplay] XML validation failed:",
                             validation.error,
                         )
+                        // Only show toast if this is the final XML (not during streaming)
+                        if (showToast) {
                             toast.error(
                                 "Diagram validation failed. Please try regenerating.",
                             )
                         }
+                    }
                 } catch (error) {
                     console.error(
                         "[ChatMessageDisplay] Error processing XML:",
@@ -636,10 +603,17 @@ export function ChatMessageDisplay({
             }
         })
 
-        // NOTE: Don't cleanup debounce timeouts here!
+        // Cleanup: clear any pending debounce timeout on unmount
-        // The cleanup runs on every re-render (when messages changes),
+        return () => {
-        // which would cancel the timeout before it fires.
+            if (debounceTimeoutRef.current) {
-        // Let the timeouts complete naturally - they're harmless if component unmounts.
+                clearTimeout(debounceTimeoutRef.current)
+                debounceTimeoutRef.current = null
+            }
+            if (editDebounceTimeoutRef.current) {
+                clearTimeout(editDebounceTimeoutRef.current)
+                editDebounceTimeoutRef.current = null
+            }
+        }
     }, [messages, handleDisplayChart, chartXML])
 
     const renderToolPart = (part: ToolPartLike) => {

components/chat-panel.tsx

@@ -76,7 +76,6 @@ interface ChatPanelProps {
 const TOOL_ERROR_STATE = "output-error" as const
 const DEBUG = process.env.NODE_ENV === "development"
 const MAX_AUTO_RETRY_COUNT = 1
-const MAX_CONTINUATION_RETRY_COUNT = 2 // Limit for truncation continuation retries
 
 /**
  * Check if auto-resubmit should happen based on tool errors.
@@ -217,8 +216,6 @@ export default function ChatPanel({
 
     // Ref to track consecutive auto-retry count (reset on user action)
     const autoRetryCountRef = useRef(0)
-    // Ref to track continuation retry count (for truncation handling)
-    const continuationRetryCountRef = useRef(0)
 
     // Ref to accumulate partial XML when output is truncated due to maxOutputTokens
     // When partialXmlRef.current.length > 0, we're in continuation mode
@@ -556,43 +553,6 @@ Continue from EXACTLY where you stopped.`,
             }
         },
         onError: (error) => {
-            // Handle server-side quota limit (429 response)
-            // AI SDK puts the full response body in error.message for non-OK responses
-            try {
-                const data = JSON.parse(error.message)
-                if (data.type === "request") {
-                    quotaManager.showQuotaLimitToast(data.used, data.limit)
-                    return
-                }
-                if (data.type === "token") {
-                    quotaManager.showTokenLimitToast(data.used, data.limit)
-                    return
-                }
-                if (data.type === "tpm") {
-                    quotaManager.showTPMLimitToast(data.limit)
-                    return
-                }
-            } catch {
-                // Not JSON, fall through to string matching for backwards compatibility
-            }
-
-            // Fallback to string matching
-            if (error.message.includes("Daily request limit")) {
-                quotaManager.showQuotaLimitToast()
-                return
-            }
-            if (error.message.includes("Daily token limit")) {
-                quotaManager.showTokenLimitToast()
-                return
-            }
-            if (
-                error.message.includes("Rate limit exceeded") ||
-                error.message.includes("tokens per minute")
-            ) {
-                quotaManager.showTPMLimitToast()
-                return
-            }
-
             // Silence access code error in console since it's handled by UI
             if (!error.message.includes("Invalid or missing access code")) {
                 console.error("Chat error:", error)
@@ -669,6 +629,22 @@ Continue from EXACTLY where you stopped.`,
 
             // DEBUG: Log finish reason to diagnose truncation
             console.log("[onFinish] finishReason:", metadata?.finishReason)
+            console.log("[onFinish] metadata:", metadata)
+
+            if (metadata) {
+                // Use Number.isFinite to guard against NaN (typeof NaN === 'number' is true)
+                const inputTokens = Number.isFinite(metadata.inputTokens)
+                    ? (metadata.inputTokens as number)
+                    : 0
+                const outputTokens = Number.isFinite(metadata.outputTokens)
+                    ? (metadata.outputTokens as number)
+                    : 0
+                const actualTokens = inputTokens + outputTokens
+                if (actualTokens > 0) {
+                    quotaManager.incrementTokenCount(actualTokens)
+                    quotaManager.incrementTPMCount(actualTokens)
+                }
+            }
         },
         sendAutomaticallyWhen: ({ messages }) => {
             const isInContinuationMode = partialXmlRef.current.length > 0
@@ -680,25 +656,15 @@ Continue from EXACTLY where you stopped.`,
             if (!shouldRetry) {
                 // No error, reset retry count and clear state
                 autoRetryCountRef.current = 0
-                continuationRetryCountRef.current = 0
                 partialXmlRef.current = ""
                 return false
             }
 
-            // Continuation mode: limited retries for truncation handling
+            // Continuation mode: unlimited retries (truncation continuation, not real errors)
+            // Server limits to 5 steps via stepCountIs(5)
             if (isInContinuationMode) {
-                if (
+                // Don't count against retry limit for continuation
-                    continuationRetryCountRef.current >=
+                // Quota checks still apply below
-                    MAX_CONTINUATION_RETRY_COUNT
-                ) {
-                    toast.error(
-                        `Continuation retry limit reached (${MAX_CONTINUATION_RETRY_COUNT}). The diagram may be too complex.`,
-                    )
-                    continuationRetryCountRef.current = 0
-                    partialXmlRef.current = ""
-                    return false
-                }
-                continuationRetryCountRef.current++
             } else {
                 // Regular error: check retry count limit
                 if (autoRetryCountRef.current >= MAX_AUTO_RETRY_COUNT) {
@@ -713,6 +679,23 @@ Continue from EXACTLY where you stopped.`,
                 autoRetryCountRef.current++
             }
 
+            // Check quota limits before auto-retry
+            const tokenLimitCheck = quotaManager.checkTokenLimit()
+            if (!tokenLimitCheck.allowed) {
+                quotaManager.showTokenLimitToast(tokenLimitCheck.used)
+                autoRetryCountRef.current = 0
+                partialXmlRef.current = ""
+                return false
+            }
+
+            const tpmCheck = quotaManager.checkTPMLimit()
+            if (!tpmCheck.allowed) {
+                quotaManager.showTPMLimitToast()
+                autoRetryCountRef.current = 0
+                partialXmlRef.current = ""
+                return false
+            }
+
             return true
         },
     })
@@ -929,6 +912,9 @@ Continue from EXACTLY where you stopped.`,
                 xmlSnapshotsRef.current.set(messageIndex, chartXml)
                 saveXmlSnapshots()
 
+                // Check all quota limits
+                if (!checkAllQuotaLimits()) return
+
                 sendChatMessage(parts, chartXml, previousXml, sessionId)
 
                 // Token count is tracked in onFinish with actual server usage
@@ -1006,7 +992,30 @@ Continue from EXACTLY where you stopped.`,
         saveXmlSnapshots()
     }
 
-    // Send chat message with headers
+    // Check all quota limits (daily requests, tokens, TPM)
+    const checkAllQuotaLimits = (): boolean => {
+        const limitCheck = quotaManager.checkDailyLimit()
+        if (!limitCheck.allowed) {
+            quotaManager.showQuotaLimitToast()
+            return false
+        }
+
+        const tokenLimitCheck = quotaManager.checkTokenLimit()
+        if (!tokenLimitCheck.allowed) {
+            quotaManager.showTokenLimitToast(tokenLimitCheck.used)
+            return false
+        }
+
+        const tpmCheck = quotaManager.checkTPMLimit()
+        if (!tpmCheck.allowed) {
+            quotaManager.showTPMLimitToast()
+            return false
+        }
+
+        return true
+    }
+
+    // Send chat message with headers and increment quota
     const sendChatMessage = (
         parts: any,
         xml: string,
@@ -1015,7 +1024,6 @@ Continue from EXACTLY where you stopped.`,
     ) => {
         // Reset all retry/continuation state on user-initiated message
         autoRetryCountRef.current = 0
-        continuationRetryCountRef.current = 0
         partialXmlRef.current = ""
 
         const config = getSelectedAIConfig()
@@ -1056,6 +1064,7 @@ Continue from EXACTLY where you stopped.`,
                 },
             },
         )
+        quotaManager.incrementRequestCount()
     }
 
     // Process files and append content to user text (handles PDF, text, and optionally images)
@@ -1143,8 +1152,13 @@ Continue from EXACTLY where you stopped.`,
             setMessages(newMessages)
         })
 
+        // Check all quota limits
+        if (!checkAllQuotaLimits()) return
+
         // Now send the message after state is guaranteed to be updated
         sendChatMessage(userParts, savedXml, previousXml, sessionId)
+
+        // Token count is tracked in onFinish with actual server usage
     }
 
     const handleEditMessage = async (messageIndex: number, newText: string) => {
@@ -1186,8 +1200,12 @@ Continue from EXACTLY where you stopped.`,
             setMessages(newMessages)
         })
 
+        // Check all quota limits
+        if (!checkAllQuotaLimits()) return
+
         // Now send the edited message after state is guaranteed to be updated
         sendChatMessage(newParts, savedXml, previousXml, sessionId)
+        // Token count is tracked in onFinish with actual server usage
     }
 
     // Collapsed view (desktop only)

components/model-config-dialog.tsx

@@ -103,40 +103,41 @@ function ProviderLogo({
     )
 }
 
-// Configuration section with title and optional action
+// Reusable validation button component
-function ConfigSection({
+function ValidationButton({
-    title,
+    status,
-    icon: Icon,
+    onClick,
-    action,
+    disabled,
-    children,
+    dict,
 }: {
-    title: string
+    status: ValidationStatus
-    icon: React.ComponentType<{ className?: string }>
+    onClick: () => void
-    action?: React.ReactNode
+    disabled: boolean
-    children: React.ReactNode
+    dict: ReturnType<typeof useDictionary>
 }) {
     return (
-        <div className="space-y-4">
+        <Button
-            <div className="flex items-center justify-between">
+            variant={status === "success" ? "outline" : "default"}
-                <div className="flex items-center gap-2">
+            size="sm"
-                    <Icon className="h-4 w-4 text-muted-foreground" />
+            onClick={onClick}
-                    <span className="text-xs font-medium text-muted-foreground uppercase tracking-wider">
+            disabled={disabled}
-                        {title}
+            className={cn(
-                    </span>
+                "h-9 px-4 min-w-[80px]",
-                </div>
+                status === "success" &&
-                {action}
+                    "text-emerald-600 border-emerald-200 dark:border-emerald-800",
-            </div>
+            )}
-            {children}
+        >
-        </div>
+            {status === "validating" ? (
-    )
+                <Loader2 className="h-4 w-4 animate-spin" />
-}
+            ) : status === "success" ? (
-
+                <>
-// Card wrapper with subtle depth
+                    <Check className="h-4 w-4 mr-1.5" />
-function ConfigCard({ children }: { children: React.ReactNode }) {
+                    {dict.modelConfig.verified}
-    return (
+                </>
-        <div className="rounded-2xl border border-border-subtle bg-surface-2/50 p-5 space-y-5">
+            ) : (
-            {children}
+                dict.modelConfig.test
-        </div>
+            )}
+        </Button>
     )
 }
 
@@ -153,6 +154,7 @@ export function ModelConfigDialog({
     const [validationStatus, setValidationStatus] =
         useState<ValidationStatus>("idle")
     const [validationError, setValidationError] = useState<string>("")
+    const [scrollState, setScrollState] = useState({ top: false, bottom: true })
     const [customModelInput, setCustomModelInput] = useState("")
     const scrollRef = useRef<HTMLDivElement>(null)
     const validationResetTimeoutRef = useRef<ReturnType<
@@ -184,6 +186,26 @@ export function ModelConfigDialog({
         (p) => p.id === selectedProviderId,
     )
 
+    // Track scroll position for gradient shadows
+    useEffect(() => {
+        const scrollEl = scrollRef.current?.querySelector(
+            "[data-radix-scroll-area-viewport]",
+        ) as HTMLElement | null
+        if (!scrollEl) return
+
+        const handleScroll = () => {
+            const { scrollTop, scrollHeight, clientHeight } = scrollEl
+            setScrollState({
+                top: scrollTop > 10,
+                bottom: scrollTop < scrollHeight - clientHeight - 10,
+            })
+        }
+
+        handleScroll() // Initial check
+        scrollEl.addEventListener("scroll", handleScroll)
+        return () => scrollEl.removeEventListener("scroll", handleScroll)
+    }, [selectedProvider])
+
     // Cleanup validation reset timeout on unmount
     useEffect(() => {
         return () => {
@@ -368,35 +390,34 @@ export function ModelConfigDialog({
 
     return (
         <Dialog open={open} onOpenChange={onOpenChange}>
-            <DialogContent className="sm:max-w-4xl h-[80vh] max-h-[800px] overflow-hidden flex flex-col gap-0 p-0">
+            <DialogContent className="sm:max-w-3xl h-[75vh] max-h-[700px] overflow-hidden flex flex-col gap-0 p-0">
-                {/* Header */}
+                <DialogHeader className="px-6 pt-6 pb-4 border-b bg-gradient-to-r from-primary/5 via-primary/3 to-transparent">
-                <DialogHeader className="px-6 pt-6 pb-4 shrink-0">
+                    <DialogTitle className="flex items-center gap-2.5 text-xl font-semibold">
-                    <DialogTitle className="flex items-center gap-3">
+                        <div className="p-1.5 rounded-lg bg-primary/10">
-                        <div className="p-2 rounded-xl bg-surface-2">
                             <Server className="h-5 w-5 text-primary" />
                         </div>
                         {dict.modelConfig?.title || "AI Model Configuration"}
                     </DialogTitle>
-                    <DialogDescription className="mt-1">
+                    <DialogDescription className="text-sm">
                         {dict.modelConfig?.description ||
                             "Configure multiple AI providers and models for your workspace"}
                     </DialogDescription>
                 </DialogHeader>
 
-                <div className="flex flex-1 min-h-0 overflow-hidden border-t border-border-subtle">
+                <div className="flex flex-1 min-h-0 overflow-hidden">
                     {/* Provider List (Left Sidebar) */}
-                    <div className="w-60 shrink-0 flex flex-col bg-surface-1/50 border-r border-border-subtle">
+                    <div className="w-56 flex-shrink-0 flex flex-col border-r bg-muted/20">
-                        <div className="px-4 py-3">
+                        <div className="px-4 py-3 border-b">
                             <span className="text-xs font-medium text-muted-foreground uppercase tracking-wider">
                                 {dict.modelConfig.providers}
                             </span>
                         </div>
 
-                        <ScrollArea className="flex-1 px-2">
+                        <ScrollArea className="flex-1">
-                            <div className="space-y-1 pb-2">
+                            <div className="p-2">
                                 {config.providers.length === 0 ? (
                                     <div className="px-3 py-8 text-center">
-                                        <div className="inline-flex items-center justify-center w-10 h-10 rounded-full bg-surface-2 mb-3">
+                                        <div className="inline-flex items-center justify-center w-10 h-10 rounded-full bg-muted mb-3">
                                             <Plus className="h-5 w-5 text-muted-foreground" />
                                         </div>
                                         <p className="text-xs text-muted-foreground">
@@ -404,7 +425,8 @@ export function ModelConfigDialog({
                                         </p>
                                     </div>
                                 ) : (
-                                    config.providers.map((provider) => (
+                                    <div className="flex flex-col gap-1">
+                                        {config.providers.map((provider) => (
                                             <button
                                                 key={provider.id}
                                                 type="button"
@@ -420,42 +442,29 @@ export function ModelConfigDialog({
                                                     setShowApiKey(false)
                                                 }}
                                                 className={cn(
-                                                "group flex items-center gap-3 px-3 py-2.5 rounded-xl w-full",
+                                                    "group flex items-center gap-3 px-3 py-2.5 rounded-lg text-left text-sm transition-all duration-150 hover:bg-accent focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2",
-                                                "text-left text-sm transition-all duration-150",
-                                                "hover:bg-interactive-hover",
-                                                "focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2",
                                                     selectedProviderId ===
                                                         provider.id &&
-                                                    "bg-surface-0 shadow-sm ring-1 ring-border-subtle",
+                                                        "bg-background shadow-sm ring-1 ring-border",
-                                            )}
-                                        >
-                                            <div
-                                                className={cn(
-                                                    "w-8 h-8 rounded-lg flex items-center justify-center",
-                                                    "bg-surface-2 transition-colors duration-150",
-                                                    selectedProviderId ===
-                                                        provider.id &&
-                                                        "bg-primary/10",
                                                 )}
                                             >
                                                 <ProviderLogo
                                                     provider={provider.provider}
                                                     className="flex-shrink-0"
                                                 />
-                                            </div>
                                                 <span className="flex-1 truncate font-medium">
                                                     {getProviderDisplayName(
                                                         provider,
                                                     )}
                                                 </span>
                                                 {provider.validated ? (
-                                                <div className="flex-shrink-0 flex items-center justify-center w-5 h-5 rounded-full bg-success-muted">
+                                                    <div className="flex-shrink-0 flex items-center justify-center w-5 h-5 rounded-full bg-emerald-500/10">
-                                                    <Check className="h-3 w-3 text-success" />
+                                                        <Check className="h-3 w-3 text-emerald-500" />
                                                     </div>
                                                 ) : (
                                                     <ChevronRight
                                                         className={cn(
-                                                        "h-4 w-4 text-muted-foreground/50 transition-transform duration-150",
+                                                            "h-4 w-4 text-muted-foreground/50 transition-transform",
                                                             selectedProviderId ===
                                                                 provider.id &&
                                                                 "translate-x-0.5",
@@ -463,19 +472,20 @@ export function ModelConfigDialog({
                                                     />
                                                 )}
                                             </button>
-                                    ))
+                                        ))}
+                                    </div>
                                 )}
                             </div>
                         </ScrollArea>
 
                         {/* Add Provider */}
-                        <div className="p-3 border-t border-border-subtle">
+                        <div className="p-2 border-t">
                             <Select
                                 onValueChange={(v) =>
                                     handleAddProvider(v as ProviderName)
                                 }
                             >
-                                <SelectTrigger className="w-full h-9 rounded-xl bg-surface-0 border-border-subtle hover:bg-interactive-hover">
+                                <SelectTrigger className="h-9 bg-background hover:bg-accent">
                                     <Plus className="h-4 w-4 mr-2 text-muted-foreground" />
                                     <SelectValue
                                         placeholder={
@@ -504,23 +514,41 @@ export function ModelConfigDialog({
                     </div>
 
                     {/* Provider Details (Right Panel) */}
-                    <div className="flex-1 min-w-0 flex flex-col overflow-hidden">
+                    <div className="flex-1 min-w-0 overflow-hidden relative">
                         {selectedProvider ? (
                             <>
-                                <ScrollArea className="flex-1" ref={scrollRef}>
+                                {/* Top gradient shadow */}
-                                    <div className="p-6 space-y-8">
+                                <div
+                                    className={cn(
+                                        "absolute top-0 left-0 right-0 h-8 bg-gradient-to-b from-background to-transparent z-10 pointer-events-none transition-opacity duration-200",
+                                        scrollState.top
+                                            ? "opacity-100"
+                                            : "opacity-0",
+                                    )}
+                                />
+                                {/* Bottom gradient shadow */}
+                                <div
+                                    className={cn(
+                                        "absolute bottom-0 left-0 right-0 h-8 bg-gradient-to-t from-background to-transparent z-10 pointer-events-none transition-opacity duration-200",
+                                        scrollState.bottom
+                                            ? "opacity-100"
+                                            : "opacity-0",
+                                    )}
+                                />
+                                <ScrollArea className="h-full" ref={scrollRef}>
+                                    <div className="p-6 space-y-6">
                                         {/* Provider Header */}
                                         <div className="flex items-center gap-3">
-                                            <div className="flex items-center justify-center w-12 h-12 rounded-xl bg-surface-2">
+                                            <div className="flex items-center justify-center w-10 h-10 rounded-lg bg-muted">
                                                 <ProviderLogo
                                                     provider={
                                                         selectedProvider.provider
                                                     }
-                                                    className="h-6 w-6"
+                                                    className="h-5 w-5"
                                                 />
                                             </div>
                                             <div className="flex-1 min-w-0">
-                                                <h3 className="font-semibold text-lg tracking-tight">
+                                                <h3 className="font-semibold text-base">
                                                     {
                                                         PROVIDER_INFO[
                                                             selectedProvider
@@ -528,7 +556,7 @@ export function ModelConfigDialog({
                                                         ].label
                                                     }
                                                 </h3>
-                                                <p className="text-sm text-muted-foreground">
+                                                <p className="text-xs text-muted-foreground">
                                                     {selectedProvider.models
                                                         .length === 0
                                                         ? dict.modelConfig
@@ -545,8 +573,8 @@ export function ModelConfigDialog({
                                                 </p>
                                             </div>
                                             {selectedProvider.validated && (
-                                                <div className="flex items-center gap-1.5 px-3 py-1.5 rounded-full bg-success-muted text-success">
+                                                <div className="flex items-center gap-1.5 px-2.5 py-1 rounded-full bg-emerald-500/10 text-emerald-600 dark:text-emerald-400">
-                                                    <Check className="h-3.5 w-3.5 animate-check-pop" />
+                                                    <Check className="h-3.5 w-3.5" />
                                                     <span className="text-xs font-medium">
                                                         {
                                                             dict.modelConfig
@@ -558,13 +586,18 @@ export function ModelConfigDialog({
                                         </div>
 
                                         {/* Configuration Section */}
-                                        <ConfigSection
+                                        <div className="space-y-4">
-                                            title={
+                                            <div className="flex items-center gap-2 text-sm font-medium text-muted-foreground">
-                                                dict.modelConfig.configuration
+                                                <Settings2 className="h-4 w-4" />
+                                                <span>
+                                                    {
+                                                        dict.modelConfig
+                                                            .configuration
                                                     }
-                                            icon={Settings2}
+                                                </span>
-                                        >
+                                            </div>
-                                            <ConfigCard>
+
+                                            <div className="rounded-xl border bg-card p-4 space-y-4">
                                                 {/* Display Name */}
                                                 <div className="space-y-2">
                                                     <Label
@@ -823,7 +856,7 @@ export function ModelConfigDialog({
                                                                     "h-9 px-4",
                                                                     validationStatus ===
                                                                         "success" &&
-                                                                        "text-success border-success/30 bg-success-muted hover:bg-success-muted",
+                                                                        "text-emerald-600 border-emerald-200 dark:border-emerald-800",
                                                                 )}
                                                             >
                                                                 {validationStatus ===
@@ -832,7 +865,7 @@ export function ModelConfigDialog({
                                                                 ) : validationStatus ===
                                                                   "success" ? (
                                                                     <>
-                                                                        <Check className="h-4 w-4 mr-1.5 animate-check-pop" />
+                                                                        <Check className="h-4 w-4 mr-1.5" />
                                                                         {
                                                                             dict
                                                                                 .modelConfig
@@ -942,7 +975,7 @@ export function ModelConfigDialog({
                                                                         "h-9 px-4",
                                                                         validationStatus ===
                                                                             "success" &&
-                                                                            "text-success border-success/30 bg-success-muted hover:bg-success-muted",
+                                                                            "text-emerald-600 border-emerald-200 dark:border-emerald-800",
                                                                     )}
                                                                 >
                                                                     {validationStatus ===
@@ -951,7 +984,7 @@ export function ModelConfigDialog({
                                                                     ) : validationStatus ===
                                                                       "success" ? (
                                                                         <>
-                                                                            <Check className="h-4 w-4 mr-1.5 animate-check-pop" />
+                                                                            <Check className="h-4 w-4 mr-1.5" />
                                                                             {
                                                                                 dict
                                                                                     .modelConfig
@@ -1020,19 +1053,26 @@ export function ModelConfigDialog({
                                                                         .modelConfig
                                                                         .customEndpoint
                                                                 }
-                                                                className="h-9 rounded-xl font-mono text-xs"
+                                                                className="h-9 font-mono text-xs"
                                                             />
                                                         </div>
                                                     </>
                                                 )}
-                                            </ConfigCard>
+                                            </div>
-                                        </ConfigSection>
+                                        </div>
 
                                         {/* Models Section */}
-                                        <ConfigSection
+                                        <div className="space-y-4">
-                                            title={dict.modelConfig.models}
+                                            <div className="flex items-center justify-between">
-                                            icon={Sparkles}
+                                                <div className="flex items-center gap-2 text-sm font-medium text-muted-foreground">
-                                            action={
+                                                    <Sparkles className="h-4 w-4" />
+                                                    <span>
+                                                        {
+                                                            dict.modelConfig
+                                                                .models
+                                                        }
+                                                    </span>
+                                                </div>
                                                 <div className="flex items-center gap-2">
                                                     <div className="relative">
                                                         <Input
@@ -1048,6 +1088,7 @@ export function ModelConfigDialog({
                                                                     e.target
                                                                         .value,
                                                                 )
+                                                                // Clear duplicate error when typing
                                                                 if (
                                                                     duplicateError
                                                                 ) {
@@ -1076,11 +1117,12 @@ export function ModelConfigDialog({
                                                                 }
                                                             }}
                                                             className={cn(
-                                                                "h-8 w-44 rounded-lg font-mono text-xs",
+                                                                "h-8 w-48 font-mono text-xs",
                                                                 duplicateError &&
                                                                     "border-destructive focus-visible:ring-destructive",
                                                             )}
                                                         />
+                                                        {/* Show duplicate error for custom model input */}
                                                         {duplicateError && (
                                                             <p className="absolute top-full left-0 mt-1 text-[11px] text-destructive">
                                                                 {duplicateError}
@@ -1090,7 +1132,7 @@ export function ModelConfigDialog({
                                                     <Button
                                                         variant="outline"
                                                         size="sm"
-                                                        className="h-8 rounded-lg"
+                                                        className="h-8"
                                                         onClick={() => {
                                                             if (
                                                                 customModelInput.trim()
@@ -1127,7 +1169,7 @@ export function ModelConfigDialog({
                                                             0
                                                         }
                                                     >
-                                                        <SelectTrigger className="w-28 h-8 rounded-lg hover:bg-interactive-hover">
+                                                        <SelectTrigger className="w-32 h-8 hover:bg-accent">
                                                             <span className="text-xs">
                                                                 {availableSuggestions.length ===
                                                                 0
@@ -1160,14 +1202,14 @@ export function ModelConfigDialog({
                                                         </SelectContent>
                                                     </Select>
                                                 </div>
-                                            }
+                                            </div>
-                                        >
+
                                             {/* Model List */}
-                                            <div className="rounded-2xl border border-border-subtle bg-surface-2/30 overflow-hidden min-h-[120px]">
+                                            <div className="rounded-xl border bg-card overflow-hidden min-h-[120px]">
                                                 {selectedProvider.models
                                                     .length === 0 ? (
-                                                    <div className="p-6 text-center h-full flex flex-col items-center justify-center">
+                                                    <div className="p-4 text-center h-full flex flex-col items-center justify-center">
-                                                        <div className="inline-flex items-center justify-center w-10 h-10 rounded-full bg-surface-2 mb-3">
+                                                        <div className="inline-flex items-center justify-center w-10 h-10 rounded-full bg-muted mb-2">
                                                             <Sparkles className="h-5 w-5 text-muted-foreground" />
                                                         </div>
                                                         <p className="text-sm text-muted-foreground">
@@ -1178,7 +1220,7 @@ export function ModelConfigDialog({
                                                         </p>
                                                     </div>
                                                 ) : (
-                                                    <div className="divide-y divide-border-subtle">
+                                                    <div className="divide-y">
                                                         {selectedProvider.models.map(
                                                             (model, index) => (
                                                                 <div
@@ -1186,7 +1228,16 @@ export function ModelConfigDialog({
                                                                         model.id
                                                                     }
                                                                     className={cn(
-                                                                        "transition-colors duration-150 hover:bg-interactive-hover/50",
+                                                                        "transition-colors hover:bg-muted/30",
+                                                                        index ===
+                                                                            0 &&
+                                                                            "rounded-t-xl",
+                                                                        index ===
+                                                                            selectedProvider
+                                                                                .models
+                                                                                .length -
+                                                                                1 &&
+                                                                            "rounded-b-xl",
                                                                     )}
                                                                 >
                                                                     <div className="flex items-center gap-3 p-3 min-w-0">
@@ -1213,8 +1264,8 @@ export function ModelConfigDialog({
                                                                             ) : model.validated ===
                                                                               true ? (
                                                                                 // Valid
-                                                                                <div className="w-full h-full rounded-lg bg-success-muted flex items-center justify-center">
+                                                                                <div className="w-full h-full rounded-lg bg-emerald-500/10 flex items-center justify-center">
-                                                                                    <Check className="h-4 w-4 text-success" />
+                                                                                    <Check className="h-4 w-4 text-emerald-500" />
                                                                                 </div>
                                                                             ) : model.validated ===
                                                                               false ? (
@@ -1415,7 +1466,7 @@ export function ModelConfigDialog({
                                                     </div>
                                                 )}
                                             </div>
-                                        </ConfigSection>
+                                        </div>
 
                                         {/* Danger Zone */}
                                         <div className="pt-4">
@@ -1425,7 +1476,7 @@ export function ModelConfigDialog({
                                                 onClick={() =>
                                                     setDeleteConfirmOpen(true)
                                                 }
-                                                className="text-muted-foreground hover:text-destructive hover:bg-destructive/5 rounded-xl"
+                                                className="text-muted-foreground hover:text-destructive hover:bg-destructive/10"
                                             >
                                                 <Trash2 className="h-4 w-4 mr-2" />
                                                 {
@@ -1439,10 +1490,10 @@ export function ModelConfigDialog({
                             </>
                         ) : (
                             <div className="h-full flex flex-col items-center justify-center p-8 text-center">
-                                <div className="inline-flex items-center justify-center w-16 h-16 rounded-2xl bg-surface-2 mb-4">
+                                <div className="inline-flex items-center justify-center w-16 h-16 rounded-2xl bg-gradient-to-br from-primary/10 to-primary/5 mb-4">
-                                    <Server className="h-8 w-8 text-muted-foreground" />
+                                    <Server className="h-8 w-8 text-primary/60" />
                                 </div>
-                                <h3 className="font-semibold text-lg tracking-tight mb-1">
+                                <h3 className="font-semibold mb-1">
                                     {dict.modelConfig.configureProviders}
                                 </h3>
                                 <p className="text-sm text-muted-foreground max-w-xs">
@@ -1454,7 +1505,7 @@ export function ModelConfigDialog({
                 </div>
 
                 {/* Footer */}
-                <div className="px-6 py-3 border-t border-border-subtle bg-surface-1/30 shrink-0">
+                <div className="px-6 py-3 border-t bg-muted/20">
                     <p className="text-xs text-muted-foreground text-center flex items-center justify-center gap-1.5">
                         <Key className="h-3 w-3" />
                         {dict.modelConfig.apiKeyStored}

components/settings-dialog.tsx

@@ -24,32 +24,6 @@ import { Switch } from "@/components/ui/switch"
 import { useDictionary } from "@/hooks/use-dictionary"
 import { getApiEndpoint } from "@/lib/base-path"
 import { i18n, type Locale } from "@/lib/i18n/config"
-import { cn } from "@/lib/utils"
-
-// Reusable setting item component for consistent layout
-function SettingItem({
-    label,
-    description,
-    children,
-}: {
-    label: string
-    description?: string
-    children: React.ReactNode
-}) {
-    return (
-        <div className="flex items-center justify-between py-4 first:pt-0 last:pb-0">
-            <div className="space-y-0.5 pr-4">
-                <Label className="text-sm font-medium">{label}</Label>
-                {description && (
-                    <p className="text-xs text-muted-foreground max-w-[260px]">
-                        {description}
-                    </p>
-                )}
-            </div>
-            <div className="shrink-0">{children}</div>
-        </div>
-    )
-}
 
 const LANGUAGE_LABELS: Record<Locale, string> = {
     en: "English",
@@ -203,76 +177,60 @@ function SettingsContent({
     }
 
     return (
-        <DialogContent className="sm:max-w-lg p-0 gap-0">
+        <DialogContent className="sm:max-w-md">
-            {/* Header */}
+            <DialogHeader>
-            <DialogHeader className="px-6 pt-6 pb-4">
                 <DialogTitle>{dict.settings.title}</DialogTitle>
-                <DialogDescription className="mt-1">
+                <DialogDescription>
                     {dict.settings.description}
                 </DialogDescription>
             </DialogHeader>
-
+            <div className="space-y-4 py-2">
-            {/* Content */}
-            <div className="px-6 pb-6">
-                <div className="divide-y divide-border-subtle">
-                    {/* Access Code (conditional) */}
                 {accessCodeRequired && (
-                        <div className="py-4 first:pt-0 space-y-3">
+                    <div className="space-y-2">
-                            <div className="space-y-0.5">
+                        <Label htmlFor="access-code">
-                                <Label
-                                    htmlFor="access-code"
-                                    className="text-sm font-medium"
-                                >
                             {dict.settings.accessCode}
                         </Label>
-                                <p className="text-xs text-muted-foreground">
-                                    {dict.settings.accessCodeDescription}
-                                </p>
-                            </div>
                         <div className="flex gap-2">
                             <Input
                                 id="access-code"
                                 type="password"
                                 value={accessCode}
-                                    onChange={(e) =>
+                                onChange={(e) => setAccessCode(e.target.value)}
-                                        setAccessCode(e.target.value)
-                                    }
                                 onKeyDown={handleKeyDown}
                                 placeholder={
                                     dict.settings.accessCodePlaceholder
                                 }
                                 autoComplete="off"
-                                    className="h-9"
                             />
                             <Button
                                 onClick={handleSave}
                                 disabled={isVerifying || !accessCode.trim()}
-                                    className="h-9 px-4 rounded-xl"
                             >
                                 {isVerifying ? "..." : dict.common.save}
                             </Button>
                         </div>
+                        <p className="text-[0.8rem] text-muted-foreground">
+                            {dict.settings.accessCodeDescription}
+                        </p>
                         {error && (
-                                <p className="text-xs text-destructive">
+                            <p className="text-[0.8rem] text-destructive">
                                 {error}
                             </p>
                         )}
                     </div>
                 )}
 
-                    {/* Language */}
+                <div className="flex items-center justify-between">
-                    <SettingItem
+                    <div className="space-y-0.5">
-                        label={dict.settings.language}
+                        <Label htmlFor="language-select">
-                        description={dict.settings.languageDescription}
+                            {dict.settings.language}
-                    >
+                        </Label>
-                        <Select
+                        <p className="text-[0.8rem] text-muted-foreground">
-                            value={currentLang}
+                            {dict.settings.languageDescription}
-                            onValueChange={changeLanguage}
+                        </p>
-                        >
+                    </div>
-                            <SelectTrigger
+                    <Select value={currentLang} onValueChange={changeLanguage}>
-                                id="language-select"
+                        <SelectTrigger id="language-select" className="w-32">
-                                className="w-[120px] h-9 rounded-xl"
-                            >
                             <SelectValue />
                         </SelectTrigger>
                         <SelectContent>
@@ -283,19 +241,22 @@ function SettingsContent({
                             ))}
                         </SelectContent>
                     </Select>
-                    </SettingItem>
+                </div>
 
-                    {/* Theme */}
+                <div className="flex items-center justify-between">
-                    <SettingItem
+                    <div className="space-y-0.5">
-                        label={dict.settings.theme}
+                        <Label htmlFor="theme-toggle">
-                        description={dict.settings.themeDescription}
+                            {dict.settings.theme}
-                    >
+                        </Label>
+                        <p className="text-[0.8rem] text-muted-foreground">
+                            {dict.settings.themeDescription}
+                        </p>
+                    </div>
                     <Button
                         id="theme-toggle"
                         variant="outline"
                         size="icon"
                         onClick={onToggleDarkMode}
-                            className="h-9 w-9 rounded-xl border-border-subtle hover:bg-interactive-hover"
                     >
                         {darkMode ? (
                             <Sun className="h-4 w-4" />
@@ -303,35 +264,42 @@ function SettingsContent({
                             <Moon className="h-4 w-4" />
                         )}
                     </Button>
-                    </SettingItem>
+                </div>
 
-                    {/* Draw.io Style */}
+                <div className="flex items-center justify-between">
-                    <SettingItem
+                    <div className="space-y-0.5">
-                        label={dict.settings.drawioStyle}
+                        <Label htmlFor="drawio-ui">
-                        description={`${dict.settings.drawioStyleDescription} ${
+                            {dict.settings.drawioStyle}
-                            drawioUi === "min"
+                        </Label>
+                        <p className="text-[0.8rem] text-muted-foreground">
+                            {dict.settings.drawioStyleDescription}{" "}
+                            {drawioUi === "min"
                                 ? dict.settings.minimal
-                                : dict.settings.sketch
+                                : dict.settings.sketch}
-                        }`}
+                        </p>
-                    >
+                    </div>
                     <Button
                         id="drawio-ui"
                         variant="outline"
+                        size="sm"
                         onClick={onToggleDrawioUi}
-                            className="h-9 w-[120px] rounded-xl border-border-subtle hover:bg-interactive-hover font-normal"
                     >
                         {dict.settings.switchTo}{" "}
                         {drawioUi === "min"
                             ? dict.settings.sketch
                             : dict.settings.minimal}
                     </Button>
-                    </SettingItem>
+                </div>
 
-                    {/* Close Protection */}
+                <div className="flex items-center justify-between">
-                    <SettingItem
+                    <div className="space-y-0.5">
-                        label={dict.settings.closeProtection}
+                        <Label htmlFor="close-protection">
-                        description={dict.settings.closeProtectionDescription}
+                            {dict.settings.closeProtection}
-                    >
+                        </Label>
+                        <p className="text-[0.8rem] text-muted-foreground">
+                            {dict.settings.closeProtectionDescription}
+                        </p>
+                    </div>
                     <Switch
                         id="close-protection"
                         checked={closeProtection}
@@ -344,13 +312,10 @@ function SettingsContent({
                             onCloseProtectionChange?.(checked)
                         }}
                     />
-                    </SettingItem>
                 </div>
             </div>
-
+            <div className="pt-4 border-t border-border/50">
-            {/* Footer */}
+                <p className="text-[0.75rem] text-muted-foreground text-center">
-            <div className="px-6 py-4 border-t border-border-subtle bg-surface-1/50 rounded-b-2xl">
-                <p className="text-xs text-muted-foreground text-center">
                     Version {process.env.APP_VERSION}
                 </p>
             </div>
@@ -363,9 +328,9 @@ export function SettingsDialog(props: SettingsDialogProps) {
         <Dialog open={props.open} onOpenChange={props.onOpenChange}>
             <Suspense
                 fallback={
-                    <DialogContent className="sm:max-w-lg p-0">
+                    <DialogContent className="sm:max-w-md">
-                        <div className="h-80 flex items-center justify-center">
+                        <div className="h-64 flex items-center justify-center">
-                            <div className="animate-spin h-6 w-6 border-2 border-primary border-t-transparent rounded-full" />
+                            <div className="animate-spin h-8 w-8 border-4 border-primary border-t-transparent rounded-full" />
                         </div>
                     </DialogContent>
                 }

components/ui/dialog.tsx

@@ -38,10 +38,7 @@ function DialogOverlay({
     <DialogPrimitive.Overlay
       data-slot="dialog-overlay"
       className={cn(
-        "fixed inset-0 z-50 bg-black/40 backdrop-blur-[2px]",
+        "data-[state=open]:animate-in data-[state=closed]:animate-out data-[state=closed]:fade-out-0 data-[state=open]:fade-in-0 fixed inset-0 z-50 bg-black/50",
-        "data-[state=open]:animate-in data-[state=closed]:animate-out",
-        "data-[state=closed]:fade-out-0 data-[state=open]:fade-in-0",
-        "duration-200",
         className
       )}
       {...props}
@@ -60,32 +57,13 @@ function DialogContent({
       <DialogPrimitive.Content
         data-slot="dialog-content"
         className={cn(
-          // Base styles
+          "bg-background data-[state=open]:animate-in data-[state=closed]:animate-out data-[state=closed]:fade-out-0 data-[state=open]:fade-in-0 data-[state=closed]:zoom-out-95 data-[state=open]:zoom-in-95 fixed top-[50%] left-[50%] z-50 grid w-full max-w-[calc(100%-2rem)] translate-x-[-50%] translate-y-[-50%] gap-4 rounded-lg border p-6 shadow-lg duration-200 sm:max-w-lg",
-          "fixed top-[50%] left-[50%] z-50 w-full",
-          "max-w-[calc(100%-2rem)] translate-x-[-50%] translate-y-[-50%]",
-          "grid gap-4 p-6",
-          // Refined visual treatment
-          "bg-surface-0 rounded-2xl border border-border-subtle shadow-dialog",
-          // Entry/exit animations
-          "data-[state=open]:animate-in data-[state=closed]:animate-out",
-          "data-[state=closed]:fade-out-0 data-[state=open]:fade-in-0",
-          "data-[state=closed]:zoom-out-[0.98] data-[state=open]:zoom-in-[0.98]",
-          "data-[state=closed]:slide-out-to-top-[2%] data-[state=open]:slide-in-from-top-[2%]",
-          "duration-200 sm:max-w-lg",
           className
         )}
         {...props}
       >
         {children}
-        <DialogPrimitive.Close className={cn(
+        <DialogPrimitive.Close className="ring-offset-background focus:ring-ring data-[state=open]:bg-accent data-[state=open]:text-muted-foreground absolute top-4 right-4 rounded-xs opacity-70 transition-opacity hover:opacity-100 focus:ring-2 focus:ring-offset-2 focus:outline-hidden disabled:pointer-events-none [&_svg]:pointer-events-none [&_svg]:shrink-0 [&_svg:not([class*='size-'])]:size-4">
-          "absolute top-4 right-4 rounded-xl p-1.5",
-          "text-muted-foreground/60 hover:text-foreground",
-          "hover:bg-interactive-hover",
-          "transition-all duration-150",
-          "focus:outline-none focus:ring-2 focus:ring-ring focus:ring-offset-2",
-          "disabled:pointer-events-none",
-          "[&_svg]:pointer-events-none [&_svg]:shrink-0 [&_svg]:size-4"
-        )}>
           <XIcon />
           <span className="sr-only">Close</span>
         </DialogPrimitive.Close>
@@ -124,10 +102,7 @@ function DialogTitle({
   return (
     <DialogPrimitive.Title
       data-slot="dialog-title"
-      className={cn(
+      className={cn("text-lg leading-none font-semibold", className)}
-        "text-xl font-semibold tracking-tight leading-tight",
-        className
-      )}
       {...props}
     />
   )
@@ -140,10 +115,7 @@ function DialogDescription({
   return (
     <DialogPrimitive.Description
       data-slot="dialog-description"
-      className={cn(
+      className={cn("text-muted-foreground text-sm", className)}
-        "text-sm text-muted-foreground leading-relaxed",
-        className
-      )}
       {...props}
     />
   )

components/ui/input.tsx

@@ -8,30 +8,9 @@ function Input({ className, type, ...props }: React.ComponentProps<"input">) {
       type={type}
       data-slot="input"
       className={cn(
-        // Base styles
+        "file:text-foreground placeholder:text-muted-foreground selection:bg-primary selection:text-primary-foreground dark:bg-input/30 border-input flex h-9 w-full min-w-0 rounded-md border bg-transparent px-3 py-1 text-base shadow-xs transition-[color,box-shadow] outline-none file:inline-flex file:h-7 file:border-0 file:bg-transparent file:text-sm file:font-medium disabled:pointer-events-none disabled:cursor-not-allowed disabled:opacity-50 md:text-sm",
-        "flex h-10 w-full min-w-0 rounded-xl px-3.5 py-2",
+        "focus-visible:border-ring focus-visible:ring-ring/50 focus-visible:ring-[3px]",
-        "border border-border-subtle bg-surface-1",
+        "aria-invalid:ring-destructive/20 dark:aria-invalid:ring-destructive/40 aria-invalid:border-destructive",
-        "text-sm text-foreground",
-        // Placeholder
-        "placeholder:text-muted-foreground/60",
-        // Selection
-        "selection:bg-primary selection:text-primary-foreground",
-        // Transitions
-        "transition-all duration-150 ease-out",
-        // Hover state
-        "hover:border-border-default",
-        // Focus state - refined ring
-        "focus:outline-none focus:border-primary focus:ring-2 focus:ring-primary/10",
-        // File input
-        "file:text-foreground file:inline-flex file:h-7 file:border-0",
-        "file:bg-transparent file:text-sm file:font-medium",
-        // Disabled
-        "disabled:pointer-events-none disabled:cursor-not-allowed disabled:opacity-50",
-        // Invalid state
-        "aria-invalid:border-destructive aria-invalid:ring-destructive/20",
-        "dark:aria-invalid:ring-destructive/40",
-        // Dark mode background
-        "dark:bg-surface-1",
         className
       )}
       {...props}

lib/ai-providers.ts

@@ -588,15 +588,13 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         case "openai": {
             const apiKey = overrides?.apiKey || process.env.OPENAI_API_KEY
             const baseURL = overrides?.baseUrl || process.env.OPENAI_BASE_URL
-            if (baseURL) {
+            if (baseURL || overrides?.apiKey) {
-                // Custom base URL = third-party proxy, use Chat Completions API
+                const customOpenAI = createOpenAI({
-                // for compatibility (most proxies don't support /responses endpoint)
+                    apiKey,
-                const customOpenAI = createOpenAI({ apiKey, baseURL })
+                    ...(baseURL && { baseURL }),
-                model = customOpenAI.chat(modelId)
+                })
-            } else if (overrides?.apiKey) {
+                // Use Responses API (default) instead of .chat() to support reasoning
-                // Custom API key but official OpenAI endpoint, use Responses API
+                // for gpt-5, o1, o3, o4 models. Chat Completions API does not emit reasoning events.
-                // to support reasoning for gpt-5, o1, o3, o4 models
-                const customOpenAI = createOpenAI({ apiKey })
                 model = customOpenAI(modelId)
             } else {
                 model = openai(modelId)

lib/dynamo-quota-manager.ts

@@ -1,238 +0,0 @@
-import {
-    ConditionalCheckFailedException,
-    DynamoDBClient,
-    GetItemCommand,
-    UpdateItemCommand,
-} from "@aws-sdk/client-dynamodb"
-
-// Quota tracking is OPT-IN: only enabled if DYNAMODB_QUOTA_TABLE is explicitly set
-// OSS users who don't need quota tracking can simply not set this env var
-const TABLE = process.env.DYNAMODB_QUOTA_TABLE
-const DYNAMODB_REGION = process.env.DYNAMODB_REGION || "ap-northeast-1"
-
-// Only create client if quota is enabled
-const client = TABLE ? new DynamoDBClient({ region: DYNAMODB_REGION }) : null
-
-/**
- * Check if server-side quota tracking is enabled.
- * Quota is opt-in: only enabled when DYNAMODB_QUOTA_TABLE env var is set.
- */
-export function isQuotaEnabled(): boolean {
-    return !!TABLE
-}
-
-interface QuotaLimits {
-    requests: number // Daily request limit
-    tokens: number // Daily token limit
-    tpm: number // Tokens per minute
-}
-
-interface QuotaCheckResult {
-    allowed: boolean
-    error?: string
-    type?: "request" | "token" | "tpm"
-    used?: number
-    limit?: number
-}
-
-/**
- * Check all quotas and increment request count atomically.
- * Uses ConditionExpression to prevent race conditions.
- * Returns which limit was exceeded if any.
- */
-export async function checkAndIncrementRequest(
-    ip: string,
-    limits: QuotaLimits,
-): Promise<QuotaCheckResult> {
-    // Skip if quota tracking not enabled
-    if (!client || !TABLE) {
-        return { allowed: true }
-    }
-
-    const today = new Date().toISOString().split("T")[0]
-    const currentMinute = Math.floor(Date.now() / 60000).toString()
-    const ttl = Math.floor(Date.now() / 1000) + 7 * 24 * 60 * 60
-
-    try {
-        // Atomic check-and-increment with ConditionExpression
-        // This prevents race conditions by failing if limits are exceeded
-        await client.send(
-            new UpdateItemCommand({
-                TableName: TABLE,
-                Key: { PK: { S: `IP#${ip}` } },
-                // Reset counts if new day/minute, then increment request count
-                UpdateExpression: `
-                    SET lastResetDate = :today,
-                        dailyReqCount = if_not_exists(dailyReqCount, :zero) + :one,
-                        dailyTokenCount = if_not_exists(dailyTokenCount, :zero),
-                        lastMinute = :minute,
-                        tpmCount = if_not_exists(tpmCount, :zero),
-                        #ttl = :ttl
-                `,
-                // Atomic condition: only succeed if ALL limits pass
-                // Uses attribute_not_exists for new items, then checks limits for existing items
-                ConditionExpression: `
-                    (attribute_not_exists(lastResetDate) OR lastResetDate < :today OR
-                     ((attribute_not_exists(dailyReqCount) OR dailyReqCount < :reqLimit) AND
-                      (attribute_not_exists(dailyTokenCount) OR dailyTokenCount < :tokenLimit))) AND
-                    (attribute_not_exists(lastMinute) OR lastMinute <> :minute OR
-                     attribute_not_exists(tpmCount) OR tpmCount < :tpmLimit)
-                `,
-                ExpressionAttributeNames: { "#ttl": "ttl" },
-                ExpressionAttributeValues: {
-                    ":today": { S: today },
-                    ":zero": { N: "0" },
-                    ":one": { N: "1" },
-                    ":minute": { S: currentMinute },
-                    ":ttl": { N: String(ttl) },
-                    ":reqLimit": { N: String(limits.requests || 999999) },
-                    ":tokenLimit": { N: String(limits.tokens || 999999) },
-                    ":tpmLimit": { N: String(limits.tpm || 999999) },
-                },
-            }),
-        )
-
-        return { allowed: true }
-    } catch (e: any) {
-        // Condition failed - need to determine which limit was exceeded
-        if (e instanceof ConditionalCheckFailedException) {
-            // Get current counts to determine which limit was hit
-            try {
-                const getResult = await client.send(
-                    new GetItemCommand({
-                        TableName: TABLE,
-                        Key: { PK: { S: `IP#${ip}` } },
-                    }),
-                )
-
-                const item = getResult.Item
-                const storedDate = item?.lastResetDate?.S
-                const storedMinute = item?.lastMinute?.S
-                const isNewDay = !storedDate || storedDate < today
-
-                const dailyReqCount = isNewDay
-                    ? 0
-                    : Number(item?.dailyReqCount?.N || 0)
-                const dailyTokenCount = isNewDay
-                    ? 0
-                    : Number(item?.dailyTokenCount?.N || 0)
-                const tpmCount =
-                    storedMinute !== currentMinute
-                        ? 0
-                        : Number(item?.tpmCount?.N || 0)
-
-                // Determine which limit was exceeded
-                if (limits.requests > 0 && dailyReqCount >= limits.requests) {
-                    return {
-                        allowed: false,
-                        type: "request",
-                        error: "Daily request limit exceeded",
-                        used: dailyReqCount,
-                        limit: limits.requests,
-                    }
-                }
-                if (limits.tokens > 0 && dailyTokenCount >= limits.tokens) {
-                    return {
-                        allowed: false,
-                        type: "token",
-                        error: "Daily token limit exceeded",
-                        used: dailyTokenCount,
-                        limit: limits.tokens,
-                    }
-                }
-                if (limits.tpm > 0 && tpmCount >= limits.tpm) {
-                    return {
-                        allowed: false,
-                        type: "tpm",
-                        error: "Rate limit exceeded (tokens per minute)",
-                        used: tpmCount,
-                        limit: limits.tpm,
-                    }
-                }
-
-                // Condition failed but no limit clearly exceeded - race condition edge case
-                // Fail safe by allowing (could be a reset race)
-                console.warn(
-                    `[quota] Condition failed but no limit exceeded for IP prefix: ${ip.slice(0, 8)}...`,
-                )
-                return { allowed: true }
-            } catch (getError: any) {
-                console.error(
-                    `[quota] Failed to get quota details after condition failure, IP prefix: ${ip.slice(0, 8)}..., error: ${getError.message}`,
-                )
-                return { allowed: true } // Fail open
-            }
-        }
-
-        // Other DynamoDB errors - fail open
-        console.error(
-            `[quota] DynamoDB error (fail-open), IP prefix: ${ip.slice(0, 8)}..., error: ${e.message}`,
-        )
-        return { allowed: true }
-    }
-}
-
-/**
- * Record token usage after response completes.
- * Uses atomic operations to update both daily token count and TPM count.
- * Handles minute boundaries atomically to prevent race conditions.
- */
-export async function recordTokenUsage(
-    ip: string,
-    tokens: number,
-): Promise<void> {
-    // Skip if quota tracking not enabled
-    if (!client || !TABLE) return
-    if (!Number.isFinite(tokens) || tokens <= 0) return
-
-    const currentMinute = Math.floor(Date.now() / 60000).toString()
-    const ttl = Math.floor(Date.now() / 1000) + 7 * 24 * 60 * 60
-
-    try {
-        // Try to update assuming same minute (most common case)
-        // Uses condition to ensure we're in the same minute
-        await client.send(
-            new UpdateItemCommand({
-                TableName: TABLE,
-                Key: { PK: { S: `IP#${ip}` } },
-                UpdateExpression:
-                    "SET #ttl = :ttl ADD dailyTokenCount :tokens, tpmCount :tokens",
-                ConditionExpression: "lastMinute = :minute",
-                ExpressionAttributeNames: { "#ttl": "ttl" },
-                ExpressionAttributeValues: {
-                    ":minute": { S: currentMinute },
-                    ":tokens": { N: String(tokens) },
-                    ":ttl": { N: String(ttl) },
-                },
-            }),
-        )
-    } catch (e: any) {
-        if (e instanceof ConditionalCheckFailedException) {
-            // Different minute - reset TPM count and set new minute
-            try {
-                await client.send(
-                    new UpdateItemCommand({
-                        TableName: TABLE,
-                        Key: { PK: { S: `IP#${ip}` } },
-                        UpdateExpression:
-                            "SET lastMinute = :minute, tpmCount = :tokens, #ttl = :ttl ADD dailyTokenCount :tokens",
-                        ExpressionAttributeNames: { "#ttl": "ttl" },
-                        ExpressionAttributeValues: {
-                            ":minute": { S: currentMinute },
-                            ":tokens": { N: String(tokens) },
-                            ":ttl": { N: String(ttl) },
-                        },
-                    }),
-                )
-            } catch (retryError: any) {
-                console.error(
-                    `[quota] Failed to record tokens (retry), IP prefix: ${ip.slice(0, 8)}..., tokens: ${tokens}, error: ${retryError.message}`,
-                )
-            }
-        } else {
-            console.error(
-                `[quota] Failed to record tokens, IP prefix: ${ip.slice(0, 8)}..., tokens: ${tokens}, error: ${e.message}`,
-            )
-        }
-    }
-}

lib/langfuse.ts

@@ -21,11 +21,9 @@ export function getLangfuseClient(): LangfuseClient | null {
     return langfuseClient
 }
 
-// Check if Langfuse is configured (both keys required)
+// Check if Langfuse is configured
 export function isLangfuseEnabled(): boolean {
-    return !!(
+    return !!process.env.LANGFUSE_PUBLIC_KEY
-        process.env.LANGFUSE_PUBLIC_KEY && process.env.LANGFUSE_SECRET_KEY
-    )
 }
 
 // Update trace with input data at the start of request
@@ -45,16 +43,34 @@ export function setTraceInput(params: {
 }
 
 // Update trace with output and end the span
-// Note: AI SDK 6 telemetry automatically reports token usage on its spans,
+export function setTraceOutput(
-// so we only need to set the output text and close our wrapper span
+    output: string,
-export function setTraceOutput(output: string) {
+    usage?: { promptTokens?: number; completionTokens?: number },
+) {
     if (!isLangfuseEnabled()) return
 
     updateActiveTrace({ output })
 
-    // End the observe() wrapper span (AI SDK creates its own child spans with usage)
     const activeSpan = api.trace.getActiveSpan()
     if (activeSpan) {
+        // Manually set usage attributes since AI SDK Bedrock streaming doesn't provide them
+        if (usage?.promptTokens) {
+            activeSpan.setAttribute("ai.usage.promptTokens", usage.promptTokens)
+            activeSpan.setAttribute(
+                "gen_ai.usage.input_tokens",
+                usage.promptTokens,
+            )
+        }
+        if (usage?.completionTokens) {
+            activeSpan.setAttribute(
+                "ai.usage.completionTokens",
+                usage.completionTokens,
+            )
+            activeSpan.setAttribute(
+                "gen_ai.usage.output_tokens",
+                usage.completionTokens,
+            )
+        }
         activeSpan.end()
     }
 }

lib/use-quota-manager.tsx

@@ -1,10 +1,11 @@
 "use client"
 
-import { useCallback } from "react"
+import { useCallback, useMemo } from "react"
 import { toast } from "sonner"
 import { QuotaLimitToast } from "@/components/quota-limit-toast"
 import { useDictionary } from "@/hooks/use-dictionary"
 import { formatMessage } from "@/lib/i18n/utils"
+import { STORAGE_KEYS } from "@/lib/storage"
 
 export interface QuotaConfig {
     dailyRequestLimit: number
@@ -12,45 +13,181 @@ export interface QuotaConfig {
     tpmLimit: number
 }
 
+export interface QuotaCheckResult {
+    allowed: boolean
+    remaining: number
+    used: number
+}
+
 /**
- * Hook for displaying quota limit toasts.
+ * Hook for managing request/token quotas and rate limiting.
- * Server-side handles actual quota enforcement via DynamoDB.
+ * Handles three types of limits:
- * This hook only provides UI feedback when limits are exceeded.
+ * - Daily request limit
+ * - Daily token limit
+ * - Tokens per minute (TPM) rate limit
+ *
+ * Users with their own API key bypass all limits.
  */
 export function useQuotaManager(config: QuotaConfig): {
-    showQuotaLimitToast: (used?: number, limit?: number) => void
+    hasOwnApiKey: () => boolean
-    showTokenLimitToast: (used?: number, limit?: number) => void
+    checkDailyLimit: () => QuotaCheckResult
-    showTPMLimitToast: (limit?: number) => void
+    checkTokenLimit: () => QuotaCheckResult
+    checkTPMLimit: () => QuotaCheckResult
+    incrementRequestCount: () => void
+    incrementTokenCount: (tokens: number) => void
+    incrementTPMCount: (tokens: number) => void
+    showQuotaLimitToast: () => void
+    showTokenLimitToast: (used: number) => void
+    showTPMLimitToast: () => void
 } {
     const { dailyRequestLimit, dailyTokenLimit, tpmLimit } = config
+
     const dict = useDictionary()
 
+    // Check if user has their own API key configured (bypass limits)
+    const hasOwnApiKey = useCallback((): boolean => {
+        const provider = localStorage.getItem(STORAGE_KEYS.aiProvider)
+        const apiKey = localStorage.getItem(STORAGE_KEYS.aiApiKey)
+        return !!(provider && apiKey)
+    }, [])
+
+    // Generic helper: Parse count from localStorage with NaN guard
+    const parseStorageCount = (key: string): number => {
+        const count = parseInt(localStorage.getItem(key) || "0", 10)
+        return Number.isNaN(count) ? 0 : count
+    }
+
+    // Generic helper: Create quota checker factory
+    const createQuotaChecker = useCallback(
+        (
+            getTimeKey: () => string,
+            timeStorageKey: string,
+            countStorageKey: string,
+            limit: number,
+        ) => {
+            return (): QuotaCheckResult => {
+                if (hasOwnApiKey())
+                    return { allowed: true, remaining: -1, used: 0 }
+                if (limit <= 0) return { allowed: true, remaining: -1, used: 0 }
+
+                const currentTime = getTimeKey()
+                const storedTime = localStorage.getItem(timeStorageKey)
+                let count = parseStorageCount(countStorageKey)
+
+                if (storedTime !== currentTime) {
+                    count = 0
+                    localStorage.setItem(timeStorageKey, currentTime)
+                    localStorage.setItem(countStorageKey, "0")
+                }
+
+                return {
+                    allowed: count < limit,
+                    remaining: limit - count,
+                    used: count,
+                }
+            }
+        },
+        [hasOwnApiKey],
+    )
+
+    // Generic helper: Create quota incrementer factory
+    const createQuotaIncrementer = useCallback(
+        (
+            getTimeKey: () => string,
+            timeStorageKey: string,
+            countStorageKey: string,
+            validateInput: boolean = false,
+        ) => {
+            return (tokens: number = 1): void => {
+                if (validateInput && (!Number.isFinite(tokens) || tokens <= 0))
+                    return
+
+                const currentTime = getTimeKey()
+                const storedTime = localStorage.getItem(timeStorageKey)
+                let count = parseStorageCount(countStorageKey)
+
+                if (storedTime !== currentTime) {
+                    count = 0
+                    localStorage.setItem(timeStorageKey, currentTime)
+                }
+
+                localStorage.setItem(countStorageKey, String(count + tokens))
+            }
+        },
+        [],
+    )
+
+    // Check daily request limit
+    const checkDailyLimit = useMemo(
+        () =>
+            createQuotaChecker(
+                () => new Date().toDateString(),
+                STORAGE_KEYS.requestDate,
+                STORAGE_KEYS.requestCount,
+                dailyRequestLimit,
+            ),
+        [createQuotaChecker, dailyRequestLimit],
+    )
+
+    // Increment request count
+    const incrementRequestCount = useMemo(
+        () =>
+            createQuotaIncrementer(
+                () => new Date().toDateString(),
+                STORAGE_KEYS.requestDate,
+                STORAGE_KEYS.requestCount,
+                false,
+            ),
+        [createQuotaIncrementer],
+    )
+
     // Show quota limit toast (request-based)
-    const showQuotaLimitToast = useCallback(
+    const showQuotaLimitToast = useCallback(() => {
-        (used?: number, limit?: number) => {
         toast.custom(
             (t) => (
                 <QuotaLimitToast
-                        used={used ?? dailyRequestLimit}
+                    used={dailyRequestLimit}
-                        limit={limit ?? dailyRequestLimit}
+                    limit={dailyRequestLimit}
                     onDismiss={() => toast.dismiss(t)}
                 />
             ),
             { duration: 15000 },
         )
-        },
+    }, [dailyRequestLimit])
-        [dailyRequestLimit],
+
+    // Check daily token limit
+    const checkTokenLimit = useMemo(
+        () =>
+            createQuotaChecker(
+                () => new Date().toDateString(),
+                STORAGE_KEYS.tokenDate,
+                STORAGE_KEYS.tokenCount,
+                dailyTokenLimit,
+            ),
+        [createQuotaChecker, dailyTokenLimit],
+    )
+
+    // Increment token count
+    const incrementTokenCount = useMemo(
+        () =>
+            createQuotaIncrementer(
+                () => new Date().toDateString(),
+                STORAGE_KEYS.tokenDate,
+                STORAGE_KEYS.tokenCount,
+                true, // Validate input tokens
+            ),
+        [createQuotaIncrementer],
     )
 
     // Show token limit toast
     const showTokenLimitToast = useCallback(
-        (used?: number, limit?: number) => {
+        (used: number) => {
             toast.custom(
                 (t) => (
                     <QuotaLimitToast
                         type="token"
-                        used={used ?? dailyTokenLimit}
+                        used={used}
-                        limit={limit ?? dailyTokenLimit}
+                        limit={dailyTokenLimit}
                         onDismiss={() => toast.dismiss(t)}
                     />
                 ),
@@ -60,24 +197,54 @@ export function useQuotaManager(config: QuotaConfig): {
         [dailyTokenLimit],
     )
 
+    // Check TPM (tokens per minute) limit
+    const checkTPMLimit = useMemo(
+        () =>
+            createQuotaChecker(
+                () => Math.floor(Date.now() / 60000).toString(),
+                STORAGE_KEYS.tpmMinute,
+                STORAGE_KEYS.tpmCount,
+                tpmLimit,
+            ),
+        [createQuotaChecker, tpmLimit],
+    )
+
+    // Increment TPM count
+    const incrementTPMCount = useMemo(
+        () =>
+            createQuotaIncrementer(
+                () => Math.floor(Date.now() / 60000).toString(),
+                STORAGE_KEYS.tpmMinute,
+                STORAGE_KEYS.tpmCount,
+                true, // Validate input tokens
+            ),
+        [createQuotaIncrementer],
+    )
+
     // Show TPM limit toast
-    const showTPMLimitToast = useCallback(
+    const showTPMLimitToast = useCallback(() => {
-        (limit?: number) => {
-            const effectiveLimit = limit ?? tpmLimit
         const limitDisplay =
-                effectiveLimit >= 1000
+            tpmLimit >= 1000 ? `${tpmLimit / 1000}k` : String(tpmLimit)
-                    ? `${effectiveLimit / 1000}k`
-                    : String(effectiveLimit)
         const message = formatMessage(dict.quota.tpmMessageDetailed, {
             limit: limitDisplay,
             seconds: 60,
         })
         toast.error(message, { duration: 8000 })
-        },
+    }, [tpmLimit, dict])
-        [tpmLimit, dict],
-    )
 
     return {
+        // Check functions
+        hasOwnApiKey,
+        checkDailyLimit,
+        checkTokenLimit,
+        checkTPMLimit,
+
+        // Increment functions
+        incrementRequestCount,
+        incrementTokenCount,
+        incrementTPMCount,
+
+        // Toast functions
         showQuotaLimitToast,
         showTokenLimitToast,
         showTPMLimitToast,

lib/utils.ts

@@ -61,47 +61,6 @@ export function isMxCellXmlComplete(xml: string | undefined | null): boolean {
     return trimmed.endsWith("/>") || trimmed.endsWith("</mxCell>")
 }
 
-/**
- * Extract only complete mxCell elements from partial/streaming XML.
- * This allows progressive rendering during streaming by ignoring incomplete trailing elements.
- * @param xml - The partial XML string (may contain incomplete trailing mxCell)
- * @returns XML string containing only complete mxCell elements
- */
-export function extractCompleteMxCells(xml: string | undefined | null): string {
-    if (!xml) return ""
-
-    const completeCells: Array<{ index: number; text: string }> = []
-
-    // Match self-closing mxCell tags: <mxCell ... />
-    // Also match mxCell with nested mxGeometry: <mxCell ...>...<mxGeometry .../></mxCell>
-    const selfClosingPattern = /<mxCell\s+[^>]*\/>/g
-    const nestedPattern = /<mxCell\s+[^>]*>[\s\S]*?<\/mxCell>/g
-
-    // Find all self-closing mxCell elements
-    let match: RegExpExecArray | null
-    while ((match = selfClosingPattern.exec(xml)) !== null) {
-        completeCells.push({ index: match.index, text: match[0] })
-    }
-
-    // Find all mxCell elements with nested content (like mxGeometry)
-    while ((match = nestedPattern.exec(xml)) !== null) {
-        completeCells.push({ index: match.index, text: match[0] })
-    }
-
-    // Sort by position to maintain order
-    completeCells.sort((a, b) => a.index - b.index)
-
-    // Remove duplicates (a self-closing match might overlap with nested match)
-    const seen = new Set<number>()
-    const uniqueCells = completeCells.filter((cell) => {
-        if (seen.has(cell.index)) return false
-        seen.add(cell.index)
-        return true
-    })
-
-    return uniqueCells.map((c) => c.text).join("\n")
-}
-
 // ============================================================================
 // XML Parsing Helpers
 // ============================================================================

package.json

@@ -24,22 +24,21 @@
         "dist:all": "npm run electron:build && npm run electron:prepare && npx electron-builder --mac --win --linux"
     },
     "dependencies": {
-        "@ai-sdk/amazon-bedrock": "^4.0.1",
+        "@ai-sdk/amazon-bedrock": "^3.0.70",
-        "@ai-sdk/anthropic": "^3.0.0",
+        "@ai-sdk/anthropic": "^2.0.44",
-        "@ai-sdk/azure": "^3.0.0",
+        "@ai-sdk/azure": "^2.0.69",
-        "@ai-sdk/deepseek": "^2.0.0",
+        "@ai-sdk/deepseek": "^1.0.30",
-        "@ai-sdk/gateway": "^3.0.0",
+        "@ai-sdk/gateway": "^2.0.21",
-        "@ai-sdk/google": "^3.0.0",
+        "@ai-sdk/google": "^2.0.0",
-        "@ai-sdk/openai": "^3.0.0",
+        "@ai-sdk/openai": "^2.0.19",
-        "@ai-sdk/react": "^3.0.1",
+        "@ai-sdk/react": "^2.0.107",
-        "@aws-sdk/client-dynamodb": "^3.957.0",
         "@aws-sdk/credential-providers": "^3.943.0",
         "@formatjs/intl-localematcher": "^0.7.2",
         "@langfuse/client": "^4.4.9",
         "@langfuse/otel": "^4.4.4",
         "@langfuse/tracing": "^4.4.9",
         "@next/third-parties": "^16.0.6",
-        "@openrouter/ai-sdk-provider": "^1.5.4",
+        "@openrouter/ai-sdk-provider": "^1.2.3",
         "@opentelemetry/exporter-trace-otlp-http": "^0.208.0",
         "@opentelemetry/sdk-trace-node": "^2.2.0",
         "@radix-ui/react-alert-dialog": "^1.1.15",
@@ -54,7 +53,7 @@
         "@radix-ui/react-tooltip": "^1.1.8",
         "@radix-ui/react-use-controllable-state": "^1.2.2",
         "@xmldom/xmldom": "^0.9.8",
-        "ai": "^6.0.1",
+        "ai": "^5.0.89",
         "base-64": "^1.0.0",
         "class-variance-authority": "^0.7.1",
         "clsx": "^2.1.1",
@@ -112,10 +111,5 @@
         "tailwindcss": "^4",
         "typescript": "^5",
         "wait-on": "^9.0.3"
-    },
-    "overrides": {
-        "@openrouter/ai-sdk-provider": {
-            "ai": "^6.0.1"
-        }
     }
 }