Automate browser based workflows with AI
Python aiapiautomationbrowserbrowser-automationcomputergptllmplaywrightpowerautomatepuppeteerpythonrpaseleniumvisionworkflow
sort pbs downloaded files (#3496)
LawyZheng7e73a55046da
diff --git a/skyvern/forge/sdk/artifact/storage/s3.py b/skyvern/forge/sdk/artifact/storage/s3.pyindex 8c705000a..2ee9b39d3 100644--- a/skyvern/forge/sdk/artifact/storage/s3.py+++ b/skyvern/forge/sdk/artifact/storage/s3.py@@ -212,8 +212,15 @@ class S3Storage(BaseStorage):file_infos: list[FileInfo] = []for key in object_keys:+ metadata = {}+ modified_at: datetime | None = None# Get metadata (including checksum)- metadata = await self.async_client.get_file_metadata(key, log_exception=False)+ try:+ object_info = await self.async_client.get_object_info(key)+ metadata = object_info.get("Metadata", {})+ modified_at = object_info.get("LastModified")+ except Exception:+ LOG.exception("Object info retrieval failed", uri=key)# Create FileInfo objectfilename = os.path.basename(key)@@ -228,6 +235,7 @@ class S3Storage(BaseStorage):url=presigned_urls[0],checksum=checksum,filename=metadata.get("original_filename", filename) if metadata else filename,+ modified_at=modified_at,)file_infos.append(file_info)diff --git a/skyvern/forge/sdk/schemas/files.py b/skyvern/forge/sdk/schemas/files.pyindex 03ccae791..eaf96946d 100644--- a/skyvern/forge/sdk/schemas/files.py+++ b/skyvern/forge/sdk/schemas/files.py@@ -1,3 +1,5 @@+from datetime import datetime+from pydantic import BaseModel, Field@@ -7,3 +9,4 @@ class FileInfo(BaseModel):url: str = Field(..., description="URL to access the file")checksum: str | None = Field(None, description="SHA-256 checksum of the file")filename: str | None = Field(None, description="Original filename")+ modified_at: datetime | None = Field(None, description="Modified time of the file")diff --git a/skyvern/webeye/schemas.py b/skyvern/webeye/schemas.pyindex d4a1449c0..faec6a4cc 100644--- a/skyvern/webeye/schemas.py+++ b/skyvern/webeye/schemas.py@@ -91,6 +91,9 @@ class BrowserSessionResponse(BaseModel):"Timeout getting downloaded files", browser_session_id=browser_session.persistent_browser_session_id)+ # Sort downloaded files by modified_at in descending order (newest first)+ downloaded_files.sort(key=lambda x: x.modified_at or datetime.min, reverse=True)+return cls(browser_session_id=browser_session.persistent_browser_session_id,organization_id=browser_session.organization_id,