M4

Instructions

Implementation Reference

• src/m4/core/tools/tabular.py:94-162 (handler)

  Core handler: GetTableInfoTool class implements the tool logic, invoking backend methods to retrieve table schema and optional sample data.
  class GetTableInfoTool: """Tool for inspecting table structure and sample data. Shows column names, data types, and optionally sample rows for a specified table. """ name = "get_table_info" description = "Get detailed information about a specific table" input_model = GetTableInfoInput output_model = ToolOutput required_modalities: frozenset[Modality] = frozenset({Modality.TABULAR}) supported_datasets: frozenset[str] | None = None def invoke( self, dataset: DatasetDefinition, params: GetTableInfoInput ) -> ToolOutput: """Get table structure and sample data using the backend.""" backend = get_backend() backend_info = backend.get_backend_info(dataset) # Validate table name if not validate_table_name(params.table_name): return ToolOutput( result=f"{backend_info}\nError: Invalid table name '{params.table_name}'" ) try: # Get table schema schema_result = backend.get_table_info(params.table_name, dataset) if not schema_result.success: return ToolOutput(result=f"{backend_info}\n{schema_result.error}") result_parts = [ backend_info, f"**Table:** {params.table_name}", "", "**Column Information:**", schema_result.data, ] # Get sample data if requested if params.show_sample: sample_result = backend.get_sample_data( params.table_name, dataset, limit=3 ) if sample_result.success: result_parts.extend( [ "", "**Sample Data (first 3 rows):**", sample_result.data, ] ) return ToolOutput(result="\n".join(result_parts)) except Exception as e: return ToolOutput( result=f"{backend_info}\nError examining table '{params.table_name}': {e}" ) def is_compatible(self, dataset: DatasetDefinition) -> bool: """Check compatibility.""" if self.supported_datasets and dataset.name not in self.supported_datasets: return False if not self.required_modalities.issubset(dataset.modalities): return False return True
• src/m4/core/tools/tabular.py:32-39 (schema)

  Input schema: Defines parameters table_name (str) and show_sample (bool) for the tool.
  @dataclass class GetTableInfoInput(ToolInput): """Input for get_table_info tool.""" table_name: str show_sample: bool = True
• src/m4/core/tools/__init__.py:65-65 (registration)

  Internal ToolRegistry registration of GetTableInfoTool instance during init_tools().
  ToolRegistry.register(GetTableInfoTool())
• src/m4/mcp_server.py:126-151 (handler)

  MCP protocol handler: Thin wrapper that registers the tool via @mcp.tool() and delegates execution to the internal ToolRegistry tool.
  @mcp.tool() @require_oauth2 def get_table_info(table_name: str, show_sample: bool = True) -> str: """🔍 Explore a specific table's structure and see sample data. **When to use:** After identifying relevant tables from get_database_schema(). Args: table_name: Exact table name (case-sensitive). show_sample: Whether to include sample rows (default: True). Returns: Table structure with column names, types, and sample data. """ dataset = DatasetRegistry.get_active() # Proactive capability check result = _tool_selector.check_compatibility("get_table_info", dataset) if not result.compatible: return result.error_message tool = ToolRegistry.get("get_table_info") return tool.invoke( dataset, GetTableInfoInput(table_name=table_name, show_sample=show_sample) ).result
• src/m4/mcp_server.py:52-52 (registration)

  MCP tool name listed in _MCP_TOOL_NAMES set for filtering and snapshots.
  "get_table_info",