Documentation
AgentOperationsResource
Access via client.agent_operations — Direct access to AI agent operations without chat sessions
synthesize()
Execute AI-powered report synthesis from images and/or documents
python
async def synthesize( intent: str, # What to synthesize *, image_ids: Optional[list[str]] = None, # Images to include document_ids: Optional[list[str]] = None, # Documents to include auto_save: bool = False, # Auto-save as document) -> SynthesizeResultReturns: SynthesizeResult - Contains report, summary, execution_time_ms, token_usage
analyze_documents()
Execute AI-powered document analysis with chunk references
python
async def analyze_documents( intent: str, # Analysis intent document_ids: list[str], # Documents to analyze (required)) -> DocumentAnalysisResultReturns: DocumentAnalysisResult - Contains analysis, summary, categorization, chunk_references
organize()
Execute AI-driven file organization. With 3+ files of a given type the system runs AI-powered clustering on image/document/video embeddings to discover natural categories, then creates folders and moves files accordingly. Below that threshold the agent reasons from the intent string alone.
python
async def organize( intent: str, # Organization intent *, image_ids: Optional[list[str]] = None, # Images to organize document_ids: Optional[list[str]] = None, # Documents to organize video_ids: Optional[list[str]] = None, # Videos to organize parent_folder_id: Optional[str] = None, # Parent folder for new folders) -> OrganizeResultReturns: OrganizeResult - Contains summary of actions taken (folders created, files moved)