Skip to main content
Glama

move_artifact

Move artifacts between folders or to root level in Mnemosyne knowledge graphs to reorganize content structure.

Instructions

Move an artifact to a different folder. Set new_parent_id to null to move to root level.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
graph_idYes
artifact_idYes
new_parent_idNo
new_orderNo

Implementation Reference

  • The handler function that executes the move_artifact tool. It authenticates, validates inputs, reads current artifact state from Y.js workspace, updates the parent folder and order using WorkspaceWriter, and returns the updated workspace snapshot.
    async def move_artifact_tool(
        graph_id: str,
        artifact_id: str,
        new_parent_id: Optional[str] = None,
        new_order: Optional[float] = None,
        context: Context | None = None,
    ) -> dict:
        """Move an artifact to a different folder via Y.js."""
        auth = MCPAuthContext.from_context(context)
        auth.require_auth()
    
        if not graph_id or not graph_id.strip():
            raise ValueError("graph_id is required and cannot be empty")
        if not artifact_id or not artifact_id.strip():
            raise ValueError("artifact_id is required and cannot be empty")
    
        try:
            await hp_client.connect_workspace(graph_id.strip())
    
            # Read current artifact state from Y.js
            channel = hp_client._workspace_channels.get(graph_id.strip())
            if channel is None:
                raise RuntimeError(f"Workspace not connected: {graph_id}")
    
            reader = WorkspaceReader(channel.doc)
            current = reader.get_artifact(artifact_id.strip())
    
            if not current:
                raise RuntimeError(f"Artifact '{artifact_id}' not found in graph '{graph_id}'")
    
            # Update artifact with new parent/order via Y.js
            await hp_client.transact_workspace(
                graph_id.strip(),
                lambda doc: WorkspaceWriter(doc).update_artifact(
                    artifact_id.strip(),
                    parent_id=new_parent_id.strip() if new_parent_id else None,
                    order=new_order,
                ),
            )
    
            snapshot = hp_client.get_workspace_snapshot(graph_id.strip())
    
            result = {
                "success": True,
                "artifact_id": artifact_id.strip(),
                "graph_id": graph_id.strip(),
                "new_parent_id": new_parent_id.strip() if new_parent_id else None,
                "workspace": snapshot,
            }
            return result
    
        except Exception as e:
            logger.error(
                "Failed to move artifact",
                extra_context={
                    "graph_id": graph_id,
                    "artifact_id": artifact_id,
                    "error": str(e),
                },
            )
            raise RuntimeError(f"Failed to move artifact: {e}")
  • The @server.tool decorator that registers the move_artifact tool with the FastMCP server, specifying its name, title, and description. The function signature provides the input schema.
    @server.tool(
        name="move_artifact",
        title="Move Artifact",
        description=(
            "Move an artifact to a different folder. "
            "Set new_parent_id to null to move to root level."
        ),
    )
  • Call to register_hocuspocus_tools which includes the move_artifact tool registration in the standalone MCP server.
    register_hocuspocus_tools(mcp_server)

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/sophia-labs/mnemosyne-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server