fix(knowledge): enforce OCR request limits by bytes and pages

Author: waleedlatif1

apps/sim/app/api/tools/mistral/parse/route.ts

@@ -9,13 +9,17 @@ import {
   validateUrlWithDNS,
 } from '@/lib/core/security/input-validation.server'
 import { generateRequestId } from '@/lib/core/utils/request'
+import { isPayloadSizeLimitError } from '@/lib/core/utils/stream-limits'
 import { withRouteHandler } from '@/lib/core/utils/with-route-handler'
 import { validateOpaqueModelInputProvenance } from '@/lib/execution/model-input-provenance'
+import { decodeDataUriWithinLimit } from '@/lib/file-parsers/data-uri'
+import { isFileParserError } from '@/lib/file-parsers/errors'
+import { MISTRAL_OCR_REQUEST_POLICY } from '@/lib/knowledge/documents/ocr-request-policy'
+import { readBoundedHttpErrorBody } from '@/lib/knowledge/documents/utils'
 import {
   isModelSafeWorkspaceFileKey,
   MODEL_UNSAFE_WORKSPACE_FILE_ERROR_MESSAGE,
 } from '@/lib/uploads/contexts/workspace/workspace-file-secret-provenance'
-import { MAX_BUFFERED_TRANSFER_BYTES } from '@/lib/uploads/shared/types'
 import {
   extractStorageKey,
   isInternalFileUrl,
@@ -160,14 +164,43 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
           requestId,
           logger,
           {
-            maxBytes: MAX_BUFFERED_TRANSFER_BYTES,
+            maxBytes: MISTRAL_OCR_REQUEST_POLICY.maxBytes,
           }
         )
         base64 = buffer.toString('base64')
         if (contentType && contentType !== 'application/octet-stream') {
           mimeType = contentType
         }
       }
+
+      let inlineBytes: number
+      try {
+        inlineBytes = base64.startsWith('data:')
+          ? decodeDataUriWithinLimit(base64, MISTRAL_OCR_REQUEST_POLICY.maxBytes).buffer.length
+          : Buffer.byteLength(base64, 'base64')
+      } catch (error) {
+        const status = isFileParserError(error) && error.code === 'complexity_limit' ? 413 : 400
+        return NextResponse.json(
+          {
+            success: false,
+            error:
+              status === 413
+                ? `File exceeds Mistral OCR's ${MISTRAL_OCR_REQUEST_POLICY.maxBytes.toLocaleString()}-byte request limit`
+                : getErrorMessage(error, 'Invalid inline file data'),
+          },
+          { status }
+        )
+      }
+      if (inlineBytes > MISTRAL_OCR_REQUEST_POLICY.maxBytes) {
+        return NextResponse.json(
+          {
+            success: false,
+            error: `File exceeds Mistral OCR's ${MISTRAL_OCR_REQUEST_POLICY.maxBytes.toLocaleString()}-byte request limit`,
+          },
+          { status: 413 }
+        )
+      }
+
       const base64Payload = base64.startsWith('data:')
         ? base64
         : `data:${mimeType};base64,${base64}`
@@ -295,8 +328,11 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
     )
 
     if (!mistralResponse.ok) {
-      const errorText = await mistralResponse.text()
-      logger.error(`[${requestId}] Mistral API error:`, errorText)
+      const errorText = await readBoundedHttpErrorBody(mistralResponse)
+      logger.error(`[${requestId}] Mistral API error`, {
+        status: mistralResponse.status,
+        diagnostic: errorText,
+      })
       return NextResponse.json(
         {
           success: false,
@@ -318,6 +354,16 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
     const notReady = docNotReadyResponse(error)
     if (notReady) return notReady
 
+    if (isPayloadSizeLimitError(error)) {
+      return NextResponse.json(
+        {
+          success: false,
+          error: `File exceeds Mistral OCR's ${MISTRAL_OCR_REQUEST_POLICY.maxBytes.toLocaleString()}-byte request limit`,
+        },
+        { status: 413 }
+      )
+    }
+
     logger.error(`[${requestId}] Error in Mistral parse:`, error)
 
     return NextResponse.json(