massgen 0.1.3__py3-none-any.whl → 0.1.4__py3-none-any.whl

massgen/configs/tools/custom_tools/multimodal_tools/text_to_speech_generation_single.yaml

@@ -0,0 +1,33 @@
+# MassGen Configuration: Text to Speech Continue Generation Tool
+# Usage:
+#    uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/text_to_speech_generation_single.yaml "I want to you tell me a very short introduction about Sherlock Homes in one sentence, and I want you to use emotion voice to read it out loud."
+agents:
+  - id: "text_to_speech_continue_generation_tool"
+    backend:
+      type: "openai"
+      model: "gpt-4o"
+      cwd: "workspace1"
+      enable_audio_generation: true
+      custom_tools:
+        - name: ["text_to_speech_transcription_generation"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/text_to_speech_transcription_generation.py"
+          function: ["text_to_speech_transcription_generation"]
+        - name: ["understand_audio"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/understand_audio.py"
+          function: ["understand_audio"]
+        - name: ["text_to_speech_continue_generation"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/text_to_speech_continue_generation.py"
+          function: ["text_to_speech_continue_generation"]
+    system_message: |
+      You are an AI assistant with access to text-to-speech generation capabilities.
+
+orchestrator:
+  snapshot_storage: "snapshots"
+  agent_temporary_workspace: "temp_workspaces"
+
+ui:
+  display_type: "simple"
+  logging_enabled: true

massgen/configs/tools/custom_tools/multimodal_tools/text_to_video_generation_multi.yaml

@@ -0,0 +1,47 @@
+# MassGen Configuration: Text to Video Generation Tool
+# Usage:
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/text_to_video_generation_multi.yaml "Generate a 4 seconds video with neon-lit alley at night, light rain, slow push-in, cinematic."
+agents:
+  - id: "text_to_video_generation_tool1"
+    backend:
+      type: "openai"
+      model: "gpt-4o"
+      cwd: "workspace1"
+      enable_video_generation: true
+      custom_tools:
+        - name: ["understand_video"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/understand_video.py"
+          function: ["understand_video"]
+        - name: ["text_to_video_generation"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/text_to_video_generation.py"
+          function: ["text_to_video_generation"]
+    system_message: |
+      You are an AI assistant with access to text-to-video generation capabilities.
+
+  - id: "text_to_video_generation_tool2"
+    backend:
+      type: "openai"
+      model: "gpt-4o"
+      cwd: "workspace2"
+      enable_video_generation: true
+      custom_tools:
+        - name: ["understand_video"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/understand_video.py"
+          function: ["understand_video"]
+        - name: ["text_to_video_generation"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/text_to_video_generation.py"
+          function: ["text_to_video_generation"]
+    system_message: |
+      You are an AI assistant with access to text-to-video generation capabilities.
+
+orchestrator:
+  snapshot_storage: "snapshots"
+  agent_temporary_workspace: "temp_workspaces"
+
+ui:
+  display_type: "rich_terminal"
+  logging_enabled: true

massgen/configs/tools/custom_tools/multimodal_tools/text_to_video_generation_single.yaml

@@ -0,0 +1,29 @@
+# MassGen Configuration: Text to Video Generation Tool
+# Usage:
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/text_to_video_generation_single.yaml "Generate a 4 seconds video with neon-lit alley at night, light rain, slow push-in, cinematic."
+agents:
+  - id: "text_to_video_generation_tool"
+    backend:
+      type: "openai"
+      model: "gpt-4o"
+      cwd: "workspace1"
+      enable_video_generation: true
+      custom_tools:
+        - name: ["understand_video"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/understand_video.py"
+          function: ["understand_video"]
+        - name: ["text_to_video_generation"]
+          category: "multimodal"
+          path: "massgen/tool/_multimodal_tools/text_to_video_generation.py"
+          function: ["text_to_video_generation"]
+    system_message: |
+      You are an AI assistant with access to text-to-video generation capabilities.
+
+orchestrator:
+  snapshot_storage: "snapshots"
+  agent_temporary_workspace: "temp_workspaces"
+
+ui:
+  display_type: "simple"
+  logging_enabled: true

massgen/configs/tools/custom_tools/multimodal_tools/understand_audio.yaml

@@ -1,6 +1,6 @@
 # MassGen Configuration: Understand Audio Tool
 # Usage:
-#   massgen --config massgen/configs/tools/custom_tools/multimodal_tools/understand_audio.yaml "Please summarize the content in this audio."
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/understand_audio.yaml "Please summarize the content in this audio."
 agents:
   - id: "understand_audio_tool"
     backend:

massgen/configs/tools/custom_tools/multimodal_tools/understand_file.yaml

@@ -1,6 +1,6 @@
 # MassGen Configuration: Understand File Tool
 # Usage:
-#   massgen --config massgen/configs/tools/custom_tools/multimodal_tools/understand_file.yaml "Please summarize the content in this file."
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/understand_file.yaml "Please summarize the content in this file."
 agents:
   - id: "understand_file_tool"
     backend:

massgen/configs/tools/custom_tools/multimodal_tools/understand_image.yaml

@@ -1,6 +1,6 @@
 # MassGen Configuration: Understand Image Tool
 # Usage:
-#   massgen --config massgen/configs/tools/custom_tools/multimodal_tools/understand_image.yaml "Please summarize the content in this image."
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/understand_image.yaml "Please summarize the content in this image."
 agents:
   - id: "understand_image_tool"
     backend:

massgen/configs/tools/custom_tools/multimodal_tools/understand_video.yaml

@@ -1,6 +1,6 @@
 # MassGen Configuration: Understand Video Tool
 # Usage:
-#   massgen --config massgen/configs/tools/custom_tools/multimodal_tools/understand_video.yaml "What's happening in this video?"
+#   uv run python -m massgen.cli --config massgen/configs/tools/custom_tools/multimodal_tools/understand_video.yaml "What's happening in this video?"
 agents:
   - id: "understand_video_tool"
     backend:

massgen/configs/tools/custom_tools/multimodal_tools/youtube_video_analysis.yaml

@@ -51,7 +51,7 @@ orchestrator:
   snapshot_storage: "snapshots"
   agent_temporary_workspace: "temp_workspaces"
   context_paths:
-    - path: "docs/case_studies"
+    - path: "docs/source/examples/case_studies"
       permission: "read"
 
 ui:

massgen/filesystem_manager/_filesystem_manager.py

@@ -57,6 +57,7 @@ class FilesystemManager:
         command_line_docker_network_mode: str = "none",
         command_line_docker_enable_sudo: bool = False,
         enable_audio_generation: bool = False,
+        enable_file_generation: bool = False,
     ):
         """
         Initialize FilesystemManager.

massgen/filesystem_manager/_path_permission_manager.py

@@ -90,6 +90,68 @@ class PathPermissionManager:
         "massgen_logs",
     ]
 
+    # Binary file extensions that should not be read by text-based tools
+    # These files should be handled by specialized tools (understand_image, understand_video, etc.)
+    BINARY_FILE_EXTENSIONS = {
+        # Images
+        ".jpg",
+        ".jpeg",
+        ".png",
+        ".gif",
+        ".bmp",
+        ".ico",
+        ".svg",
+        ".webp",
+        ".tiff",
+        ".tif",
+        # Videos
+        ".mp4",
+        ".avi",
+        ".mov",
+        ".mkv",
+        ".flv",
+        ".wmv",
+        ".webm",
+        ".m4v",
+        ".mpg",
+        ".mpeg",
+        # Audio
+        ".mp3",
+        ".wav",
+        ".ogg",
+        ".flac",
+        ".aac",
+        ".m4a",
+        ".wma",
+        # Archives
+        ".zip",
+        ".tar",
+        ".gz",
+        ".bz2",
+        ".7z",
+        ".rar",
+        ".xz",
+        # Executables and binaries
+        ".exe",
+        ".bin",
+        ".dll",
+        ".so",
+        ".dylib",
+        ".o",
+        ".a",
+        ".pyc",
+        ".class",
+        ".jar",
+        # Office documents (binary formats - use understand_file tool)
+        ".doc",  # Old Word (not supported by understand_file)
+        ".xls",  # Old Excel (not supported by understand_file)
+        ".ppt",  # Old PowerPoint (not supported by understand_file)
+        ".pdf",  # PDF (supported by understand_file with PyPDF2)
+        ".docx",  # Word (supported by understand_file with python-docx)
+        ".xlsx",  # Excel (supported by understand_file with openpyxl)
+        ".pptx",  # PowerPoint (supported by understand_file with python-pptx)
+    }
+
     def __init__(
         self,
         context_write_access_enabled: bool = False,
@@ -440,6 +502,12 @@ class PathPermissionManager:
             - allowed: Whether the tool call should proceed
             - reason: Explanation if blocked (None if allowed)
         """
+        # Check if read tool is trying to read binary files (images, videos, etc.)
+        if self._is_text_read_tool(tool_name):
+            binary_check_result = self._validate_binary_file_access(tool_name, tool_args)
+            if not binary_check_result[0]:
+                return binary_check_result
+
         # Track read operations for read-before-delete enforcement
         if self._is_read_tool(tool_name):
             self._track_read_operation(tool_name, tool_args)
@@ -495,6 +563,33 @@ class PathPermissionManager:
 
         return False
 
+    def _is_text_read_tool(self, tool_name: str) -> bool:
+        """
+        Check if a tool is a text-based read operation that should not access binary files.
+
+        These tools are designed for reading text files and should be blocked from
+        reading binary files (images, videos, audio, etc.) to prevent context pollution.
+
+        Tools that read text file contents:
+        - Read: Claude Code read tool
+        - read_text_file: MCP filesystem read tool
+        - read_file: Generic read operations
+        """
+        # Use lowercase for case-insensitive matching
+        tool_lower = tool_name.lower()
+
+        # Check if tool name contains any text read operation keywords
+        text_read_keywords = [
+            "read_text_file",  # MCP filesystem: read_text_file
+            "read_file",  # Generic read operations
+        ]
+
+        # Also check for exact "Read" match (Claude Code tool)
+        if tool_name == "Read":
+            return True
+
+        return any(keyword in tool_lower for keyword in text_read_keywords)
+
     def _is_read_tool(self, tool_name: str) -> bool:
         """
         Check if a tool is a read operation that should be tracked.
@@ -518,6 +613,59 @@ class PathPermissionManager:
 
         return any(keyword in tool_lower for keyword in read_keywords)
 
+    def _validate_binary_file_access(self, tool_name: str, tool_args: Dict[str, Any]) -> Tuple[bool, Optional[str]]:
+        """
+        Validate that text-based read tools are not trying to read binary files.
+
+        Binary files (images, videos, audio, etc.) should be handled by specialized tools
+        to prevent context pollution with binary data.
+
+        Args:
+            tool_name: Name of the tool being called
+            tool_args: Arguments passed to the tool
+
+        Returns:
+            Tuple of (allowed: bool, reason: Optional[str])
+            - allowed: False if trying to read binary file, True otherwise
+            - reason: Explanation if blocked (None if allowed)
+        """
+        # Extract file path from arguments
+        file_path = self._extract_file_path(tool_args)
+        if not file_path:
+            # Can't determine path - allow (tool may not access files)
+            return (True, None)
+
+        # Resolve path
+        try:
+            file_path_str = self._resolve_path_against_workspace(file_path)
+            path = Path(file_path_str)
+        except Exception:
+            # If path resolution fails, allow (will fail elsewhere if invalid)
+            return (True, None)
+
+        # Check file extension
+        file_extension = path.suffix.lower()
+        if file_extension in self.BINARY_FILE_EXTENSIONS:
+            # Determine appropriate tool suggestion based on file type
+            if file_extension in {".jpg", ".jpeg", ".png", ".gif", ".bmp", ".ico", ".svg", ".webp", ".tiff", ".tif"}:
+                suggestion = "For images, use understand_image tool"
+            elif file_extension in {".mp4", ".avi", ".mov", ".mkv", ".flv", ".wmv", ".webm", ".m4v", ".mpg", ".mpeg"}:
+                suggestion = "For videos, use understand_video tool"
+            elif file_extension in {".mp3", ".wav", ".ogg", ".flac", ".aac", ".m4a", ".wma"}:
+                suggestion = "For audio files, use generate_text_with_input_audio tool"
+            elif file_extension in {".pdf"}:
+                suggestion = "For PDF files, use understand_file tool"
+            elif file_extension in {".docx", ".xlsx", ".pptx"}:
+                suggestion = "For Office documents, use understand_file tool"
+            else:
+                suggestion = "Use appropriate specialized tool for this file type"
+
+            reason = f"Cannot read binary file '{path.name}' with {tool_name}. {suggestion}."
+            logger.warning(f"[PathPermissionManager] Blocked {tool_name} from reading binary file: {path}")
+            return (False, reason)
+
+        return (True, None)
+
     def _is_delete_tool(self, tool_name: str) -> bool:
         """
         Check if a tool is a delete operation.

massgen/message_templates.py

@@ -302,6 +302,8 @@ IMPORTANT: You are responding to the latest message in an ongoing conversation.
         original_system_message: Optional[str] = None,
         enable_image_generation: bool = False,
         enable_audio_generation: bool = False,
+        enable_file_generation: bool = False,
+        enable_video_generation: bool = False,
         has_irreversible_actions: bool = False,
         enable_command_execution: bool = False,
     ) -> str:
@@ -311,6 +313,8 @@ IMPORTANT: You are responding to the latest message in an ongoing conversation.
             original_system_message: The agent's original system message to preserve
             enable_image_generation: Whether image generation is enabled
             enable_audio_generation: Whether audio generation is enabled
+            enable_file_generation: Whether file generation is enabled
+            enable_video_generation: Whether video generation is enabled
             has_irreversible_actions: Whether agent has write access to context paths (requires actual file delivery)
             enable_command_execution: Whether command execution is enabled for this agent
         """
@@ -335,21 +339,165 @@ Present the best possible coordinated answer by combining the strengths from all
         # Add image generation instructions only if enabled
         if enable_image_generation:
             presentation_instructions += """For image generation tasks:
-- Extract image paths from the existing answer and resolve them in the shared reference.
-- Gather all agent-produced images (ignore non-existent files).
-- MUST call the generate-image tool with these input images to synthesize one final image combining their strengths.
-- MUST save the final outputand output the saved path.
-"""
+
+  **MANDATORY WORKFLOW - You MUST follow these steps in order:**
+
+  Step 1: **Check for existing images (REQUIRED)**
+  - First, list all files in the Shared Reference directory (temp_workspaces) to find ALL images from EVERY agent
+  - Look for image files (.png, .jpg, .jpeg, .gif, .webp, etc.) in each agent's workspace subdirectory
+
+  Step 2: **Understand ALL existing images (REQUIRED if images exist)**
+  - For EACH image file you found, you MUST call the **understand_image** tool to extract its key visual elements, composition, style, and quality
+  - Do this for images from yourself AND from other agents - analyze ALL images found
+  - DO NOT skip this step even if you think you know the content
+
+  Step 3: **Synthesize and generate final image (REQUIRED)**
+  - If existing images were found and analyzed:
+    * Synthesize ALL image analyses into a single, detailed, combined prompt
+    * The combined prompt should capture the best visual elements, composition, style, and quality from all analyzed images
+    * Call **image_to_image_generation** with this synthesized prompt and ALL images to create the final unified image
+  - If NO existing images were found:
+    * Generate a new image based directly on the original task requirements
+    * Call **text_to_image_generation** with a prompt derived from the original task
+
+  Step 4: **Save and report (REQUIRED)**
+  - Save the final generated image in your workspace
+  - Report the saved path in your final answer
+
+  **CRITICAL**: You MUST complete Steps 1-4 in order. Do not skip checking for existing images. Do not skip calling
+  understand_image on found images. This is a mandatory synthesis workflow.
+  """
+        #             presentation_instructions += """For image generation tasks:
+        # - Extract image paths from the existing answer and resolve them in the shared reference.
+        # - Gather all agent-produced images (ignore non-existent files).
+        # - IMPORTANT: If you find ANY existing images (from yourself or other agents), you MUST call the understand_image tool
+        #   to analyze EACH image and extract their key visual elements, composition, style, and quality.
+        # - IMPORTANT: Synthesize insights from all analyzed images into a detailed, combined prompt that captures the best elements.
+        # - IMPORTANT: Call text_to_image_generation with this synthesized prompt to generate the final image.
+        # - IMPORTANT: Save the final output in your workspace and output the saved path.
+        # - If no existing images are found, generate based on the original task requirements.
+        # """
         # Add audio generation instructions only if enabled
         if enable_audio_generation:
             presentation_instructions += """For audio generation tasks:
-- Extract audio paths from the existing answer and resolve them in the shared reference.
-- Gather ALL audio files produced by EVERY agent (ignore non-existent files).
-  IMPORTANT: You MUST call the generate_text_with_input_audio tool to obtain transcriptions
-  for EACH AND EVERY audio file from ALL agents - no audio should be skipped or overlooked.
-- MUST combine the strengths of all transcriptions into one final detailed transcription that captures the best elements from each.
-- MUST use the convert_text_to_audio tool to convert this final transcription to a new audio file and save it, then output the saved path.
-"""
+
+  **MANDATORY WORKFLOW - You MUST follow these steps in order:**
+
+  Step 1: **Check for existing audios (REQUIRED)**
+  - First, list all files in the Shared Reference directory (temp_workspaces) to find ALL audio files from EVERY agent
+  - Look for audio files (.mp3, .wav, .flac, etc.) in each agent's workspace subdirectory
+
+  Step 2: **Understand ALL existing audios (REQUIRED if audios exist)**
+  - For EACH audio file you found, you MUST call the **understand_audio** tool to extract its transcription
+  - Do this for audios from yourself AND from other agents - analyze ALL audios found
+  - DO NOT skip this step even if you think you know the content
+
+  Step 3: **Synthesize and generate final audio (REQUIRED)**
+  - If existing audios were found and analyzed:
+    * Synthesize ALL audio transcriptions into a single, detailed, combined transcription
+    * The combined transcription should capture the best content from all analyzed audios
+    * Call **text_to_speech_transcription_generation** with this synthesized transcription to create the final unified audio
+  - If NO existing audios were found:
+    * Generate a new audio based directly on the original task requirements
+    * Call **text_to_speech_transcription_generation** with a transcription derived from the original task
+
+  Step 4: **Save and report (REQUIRED)**
+  - Save the final generated audio in your workspace
+  - Report the saved path in your final answer
+
+  **CRITICAL**: You MUST complete Steps 1-4 in order. Do not skip checking for existing audios. Do not skip calling
+  understand_audio on found audios. This is a mandatory synthesis workflow.
+  """
+        #                         presentation_instructions += """For audio generation tasks:
+        # - Extract audio paths from the existing answer and resolve them in the shared reference.
+        # - Gather ALL audio files produced by EVERY agent (ignore non-existent files).
+        # - IMPORTANT: If you find ANY existing audios (from yourself or other agents), you MUST call the **understand_audio** tool to extract each audio's transcription.
+        # - IMPORTANT: Synthesize transcriptions from all audios into a detailed, combined transcription.
+        # - IMPORTANT: You MUST call the **text_to_speech_transcription_generation** tool with this synthesized transcription to generate the final audio.
+        # - IMPORTANT: Save the final output in your workspace and output the saved path.
+        # - If no existing audios are found, generate based on the original task requirements.
+        # """
+        # Add file generation instructions only if enabled
+        if enable_file_generation:
+            presentation_instructions += """For file generation tasks:
+
+  **MANDATORY WORKFLOW - You MUST follow these steps in order:**
+
+  Step 1: **Check for existing files (REQUIRED)**
+  - First, list all files in the Shared Reference directory (temp_workspaces) to find ALL files from EVERY agent
+  - Look for files of the requested type in each agent's workspace subdirectory
+
+  Step 2: **Understand ALL existing files (REQUIRED if files exist)**
+  - For EACH file you found, you MUST call the **understand_file** tool to extract its content, structure, and key elements
+  - Do this for files from yourself AND from other agents - analyze ALL files found
+  - DO NOT skip this step even if you think you know the content
+
+  Step 3: **Synthesize and generate final file (REQUIRED)**
+  - If existing files were found and analyzed:
+    * Synthesize ALL file contents into a single, detailed, combined content
+    * The combined content should capture the best elements, structure, and information from all analyzed files
+    * Call **text_to_file_generation** with this synthesized content to generate the final unified file
+  - If NO existing files were found:
+    * Generate a new file based directly on the original task requirements
+    * Call **text_to_file_generation** with content derived from the original task
+
+  Step 4: **Save and report (REQUIRED)**
+  - Save the final generated file in your workspace
+  - Report the saved path in your final answer
+
+  **CRITICAL**: You MUST complete Steps 1-4 in order. Do not skip checking for existing files. Do not skip calling
+  understand_file on found files. This is a mandatory synthesis workflow.
+  """
+        #             presentation_instructions += """For file generation tasks:
+        # - Extract file paths from the existing answer and resolve them in the shared reference.
+        # - Gather ALL files produced by EVERY agent (ignore non-existent files).
+        # - IMPORTANT: If you find ANY existing files (from yourself or other agents), you MUST call the **understand_file** tool to extract each file's content.
+        # - IMPORTANT: Synthesize contents from all files into a detailed, combined content.
+        # - IMPORTANT: You MUST call the **text_to_file_generation** tool with this synthesized content to generate the final file.
+        # - IMPORTANT: Save the final output in your workspace and output the saved path.
+        # - If no existing files are found, generate based on the original task requirements.
+        # """
+        # Add video generation instructions only if enabled
+        if enable_video_generation:
+            presentation_instructions += """For video generation tasks:
+
+  **MANDATORY WORKFLOW - You MUST follow these steps in order:**
+
+  Step 1: **Check for existing videos (REQUIRED)**
+  - First, list all files in the Shared Reference directory (temp_workspaces) to find ALL videos from EVERY agent
+  - Look for video files (.mp4, .avi, .mov, etc.) in each agent's workspace subdirectory
+
+  Step 2: **Understand ALL existing videos (REQUIRED if videos exist)**
+  - For EACH video file you found, you MUST call the **understand_video** tool to extract its description, visual features, and
+  key elements
+  - Do this for videos from yourself AND from other agents - analyze ALL videos found
+  - DO NOT skip this step even if you think you know the content
+
+  Step 3: **Synthesize and generate final video (REQUIRED)**
+  - If existing videos were found and analyzed:
+    * Synthesize ALL video descriptions into a single, detailed, combined prompt
+    * The combined prompt should capture the best visual elements, composition, motion, and style from all analyzed videos
+    * Call **text_to_video_generation** with this synthesized prompt to create the final unified video
+  - If NO existing videos were found:
+    * Generate a new video based directly on the original task requirements
+    * Call **text_to_video_generation** with a prompt derived from the original task
+
+  Step 4: **Save and report (REQUIRED)**
+  - Save the final generated video in your workspace
+  - Report the saved path in your final answer
+
+  **CRITICAL**: You MUST complete Steps 1-4 in order. Do not skip checking for existing videos. Do not skip calling
+  understand_video on found videos. This is a mandatory synthesis workflow.
+  """
+        #             presentation_instructions += """For video generation tasks:
+        # - Extract video paths from the existing answer and resolve them in the shared reference.
+        # - Gather ALL videos produced by EVERY agent (ignore non-existent files).
+        # - IMPORTANT: If you find ANY existing videos (from yourself or other agents), you MUST call the **understand_video** tool to extract each video's description and key features.
+        # - IMPORTANT: Synthesize descriptions from all videos into a detailed, combined prompt capturing the best elements.
+        # - IMPORTANT: You MUST call the **text_to_video_generation** tool with this synthesized prompt to generate the final video.
+        # - IMPORTANT: Save the final output in your workspace and output the saved path.
+        # - If no existing videos are found, generate based on the original task requirements.
+        # """
 
         # Add irreversible actions reminder if needed
         # TODO: Integrate more general irreversible actions handling in future (i.e., not just for context file delivery)

massgen/orchestrator.py

@@ -2513,6 +2513,20 @@ INSTRUCTIONS FOR NEXT ATTEMPT:
         elif hasattr(agent, "backend") and hasattr(agent.backend, "backend_params"):
             enable_audio_generation = agent.backend.backend_params.get("enable_audio_generation", False)
 
+        # Check if file generation is enabled for this agent
+        enable_file_generation = False
+        if hasattr(agent, "config") and agent.config:
+            enable_file_generation = agent.config.backend_params.get("enable_file_generation", False)
+        elif hasattr(agent, "backend") and hasattr(agent.backend, "backend_params"):
+            enable_file_generation = agent.backend.backend_params.get("enable_file_generation", False)
+
+        # Check if video generation is enabled for this agent
+        enable_video_generation = False
+        if hasattr(agent, "config") and agent.config:
+            enable_video_generation = agent.config.backend_params.get("enable_video_generation", False)
+        elif hasattr(agent, "backend") and hasattr(agent.backend, "backend_params"):
+            enable_video_generation = agent.backend.backend_params.get("enable_video_generation", False)
+
         # Check if agent has write access to context paths (requires file delivery)
         has_irreversible_actions = False
         if agent.backend.filesystem_manager:
@@ -2525,6 +2539,8 @@ INSTRUCTIONS FOR NEXT ATTEMPT:
             agent_system_message,
             enable_image_generation,
             enable_audio_generation,
+            enable_file_generation,
+            enable_video_generation,
             has_irreversible_actions,
             enable_command_execution,
         )