From ed26d81c2ba82f5f3d33a6d84f2429297801a594 Mon Sep 17 00:00:00 2001 From: Nicholas Tindle Date: Wed, 4 Feb 2026 18:24:43 -0600 Subject: [PATCH] fix(backend): strip MIME fragment from workspace URI before resolving file ID store_media_file() appends #mimeType fragments to workspace URIs on output (e.g. workspace://abc123#video/mp4). When a downstream block receives this URI as input, the same function was using "abc123#video/mp4" as a file ID, which fails. Add parse_workspace_uri() utility to cleanly separate the file ID from the MIME fragment, fixing block-to-block file passing. Co-Authored-By: Claude Opus 4.5 --- autogpt_platform/backend/backend/util/file.py | 64 +++++++++++++++---- 1 file changed, 51 insertions(+), 13 deletions(-) diff --git a/autogpt_platform/backend/backend/util/file.py b/autogpt_platform/backend/backend/util/file.py index baa9225629..14622b9d54 100644 --- a/autogpt_platform/backend/backend/util/file.py +++ b/autogpt_platform/backend/backend/util/file.py @@ -8,6 +8,8 @@ from pathlib import Path from typing import TYPE_CHECKING, Literal from urllib.parse import urlparse +from pydantic import BaseModel + from backend.util.cloud_storage import get_cloud_storage_handler from backend.util.request import Requests from backend.util.settings import Config @@ -17,6 +19,35 @@ from backend.util.virus_scanner import scan_content_safe if TYPE_CHECKING: from backend.data.execution import ExecutionContext + +class WorkspaceUri(BaseModel): + """Parsed workspace:// URI.""" + + file_ref: str # File ID or path (e.g. "abc123" or "/path/to/file.txt") + mime_type: str | None = None # MIME type from fragment (e.g. "video/mp4") + is_path: bool = False # True if file_ref is a path (starts with "/") + + +def parse_workspace_uri(uri: str) -> WorkspaceUri: + """Parse a workspace:// URI into its components. + + Examples: + "workspace://abc123" → WorkspaceUri(file_ref="abc123", mime_type=None, is_path=False) + "workspace://abc123#video/mp4" → WorkspaceUri(file_ref="abc123", mime_type="video/mp4", is_path=False) + "workspace:///path/to/file.txt" → WorkspaceUri(file_ref="/path/to/file.txt", mime_type=None, is_path=True) + """ + raw = uri.removeprefix("workspace://") + mime_type: str | None = None + if "#" in raw: + raw, fragment = raw.split("#", 1) + mime_type = fragment or None + return WorkspaceUri( + file_ref=raw, + mime_type=mime_type, + is_path=raw.startswith("/"), + ) + + # Return format options for store_media_file # - "for_local_processing": Returns local file path - use with ffmpeg, MoviePy, PIL, etc. # - "for_external_api": Returns data URI (base64) - use when sending content to external APIs @@ -183,22 +214,20 @@ async def store_media_file( "This file type is only available in CoPilot sessions." ) - # Parse workspace reference - # workspace://abc123 - by file ID - # workspace:///path/to/file.txt - by virtual path - file_ref = file[12:] # Remove "workspace://" + # Parse workspace reference (strips #mimeType fragment from file ID) + ws = parse_workspace_uri(file) - if file_ref.startswith("/"): - # Path reference - workspace_content = await workspace_manager.read_file(file_ref) - file_info = await workspace_manager.get_file_info_by_path(file_ref) + if ws.is_path: + # Path reference: workspace:///path/to/file.txt + workspace_content = await workspace_manager.read_file(ws.file_ref) + file_info = await workspace_manager.get_file_info_by_path(ws.file_ref) filename = sanitize_filename( file_info.name if file_info else f"{uuid.uuid4()}.bin" ) else: - # ID reference - workspace_content = await workspace_manager.read_file_by_id(file_ref) - file_info = await workspace_manager.get_file_info(file_ref) + # ID reference: workspace://abc123 or workspace://abc123#video/mp4 + workspace_content = await workspace_manager.read_file_by_id(ws.file_ref) + file_info = await workspace_manager.get_file_info(ws.file_ref) filename = sanitize_filename( file_info.name if file_info else f"{uuid.uuid4()}.bin" ) @@ -334,7 +363,16 @@ async def store_media_file( # Don't re-save if input was already from workspace if is_from_workspace: - # Return original workspace reference + # Return original workspace reference, ensuring MIME type fragment + ws = parse_workspace_uri(file) + if not ws.mime_type: + # Add MIME type fragment if missing (older refs without it) + try: + info = await workspace_manager.get_file_info(ws.file_ref) + if info: + return MediaFileType(f"{file}#{info.mimeType}") + except Exception: + pass return MediaFileType(file) # Save new content to workspace @@ -346,7 +384,7 @@ async def store_media_file( filename=filename, overwrite=True, ) - return MediaFileType(f"workspace://{file_record.id}") + return MediaFileType(f"workspace://{file_record.id}#{file_record.mimeType}") else: raise ValueError(f"Invalid return_format: {return_format}")