fix(files): deduplicate BINARY_DOC_TASKS and add size guard to VFS compiled-check

Author: waleedlatif1

apps/sim/app/api/workspaces/[id]/files/[fileId]/compiled-check/route.ts

@@ -3,23 +3,16 @@ import { toError } from '@sim/utils/errors'
 import { type NextRequest, NextResponse } from 'next/server'
 import { getSession } from '@/lib/auth'
 import { withRouteHandler } from '@/lib/core/utils/with-route-handler'
-import { MAX_DOCUMENT_PREVIEW_CODE_BYTES } from '@/lib/execution/constants'
+import { BINARY_DOC_TASKS, MAX_DOCUMENT_PREVIEW_CODE_BYTES } from '@/lib/execution/constants'
 import { runSandboxTask, SandboxUserCodeError } from '@/lib/execution/sandbox/run-task'
 import { downloadWorkspaceFile, getWorkspaceFile } from '@/lib/uploads/contexts/workspace'
 import { verifyWorkspaceMembership } from '@/app/api/workflows/utils'
-import type { SandboxTaskId } from '@/sandbox-tasks/registry'
 
 export const dynamic = 'force-dynamic'
 export const runtime = 'nodejs'
 
 const logger = createLogger('WorkspaceFileCompiledCheckAPI')
 
-const BINARY_DOC_TASKS: Record<string, SandboxTaskId> = {
-  docx: 'docx-generate',
-  pptx: 'pptx-generate',
-  pdf: 'pdf-generate',
-}
-
 /**
  * GET /api/workspaces/[id]/files/[fileId]/compiled-check
  *

apps/sim/lib/copilot/vfs/workspace-vfs.ts

@@ -58,6 +58,7 @@ import {
   getAccessibleOAuthCredentials,
 } from '@/lib/credentials/environment'
 import { getPersonalAndWorkspaceEnv } from '@/lib/environment/utils'
+import { BINARY_DOC_TASKS, MAX_DOCUMENT_PREVIEW_CODE_BYTES } from '@/lib/execution/constants'
 import { runSandboxTask, SandboxUserCodeError } from '@/lib/execution/sandbox/run-task'
 import { getKnowledgeBases } from '@/lib/knowledge/service'
 import { listTables } from '@/lib/table/service'
@@ -80,19 +81,12 @@ import {
 } from '@/lib/workspaces/permissions/utils'
 import { getAllBlocks } from '@/blocks/registry'
 import { CONNECTOR_REGISTRY } from '@/connectors/registry'
-import type { SandboxTaskId } from '@/sandbox-tasks/registry'
 import { tools as toolRegistry } from '@/tools/registry'
 import { getLatestVersionTools, stripVersionSuffix } from '@/tools/utils'
 import { TRIGGER_REGISTRY } from '@/triggers/registry'
 
 const logger = createLogger('WorkspaceVFS')
 
-const BINARY_DOC_TASKS: Record<string, SandboxTaskId> = {
-  docx: 'docx-generate',
-  pptx: 'pptx-generate',
-  pdf: 'pdf-generate',
-}
-
 /** Static component files, computed once and shared across all VFS instances */
 let staticComponentFiles: Map<string, string> | null = null
 
@@ -479,6 +473,12 @@ export class WorkspaceVFS {
         if (!taskId) return null
         const buffer = await downloadWorkspaceFile(record)
         const code = buffer.toString('utf-8')
+        if (Buffer.byteLength(code, 'utf-8') > MAX_DOCUMENT_PREVIEW_CODE_BYTES) {
+          return {
+            content: JSON.stringify({ ok: false, error: 'File source exceeds maximum size' }),
+            totalLines: 1,
+          }
+        }
         let result: { ok: boolean; error?: string; errorName?: string }
         try {
           await runSandboxTask(taskId, { code, workspaceId: this._workspaceId })

apps/sim/lib/execution/constants.ts

@@ -1,4 +1,5 @@
 import { DEFAULT_EXECUTION_TIMEOUT_MS } from '@/lib/core/execution-limits'
+import type { SandboxTaskId } from '@/sandbox-tasks/registry'
 
 export { DEFAULT_EXECUTION_TIMEOUT_MS }
 
@@ -11,3 +12,10 @@ export { DEFAULT_EXECUTION_TIMEOUT_MS }
  * while reducing memory pressure and abuse potential from oversized payloads.
  */
 export const MAX_DOCUMENT_PREVIEW_CODE_BYTES = 1 * 1024 * 1024
+
+/** Maps file extension to the sandbox task that compiles/generates that document type. */
+export const BINARY_DOC_TASKS: Record<string, SandboxTaskId> = {
+  docx: 'docx-generate',
+  pptx: 'pptx-generate',
+  pdf: 'pdf-generate',
+}