TM1 cube tools

Explore cube structures, dimensions, members, and data with natural-language or structured queries.

This tool set contains Model Context Protocol (MCP) tools for exploring the data and metadata in TM1® models. Each tool includes input arguments, outputs, and example usage in natural language.

Overview

These tools are accessible through the /agentic-ai/cube/mcp endpoint.

For REST-based authorization, this should follow the pattern https://{{PAW_HOST}}/api/{{PAW_TENANT_ID}}/v0/agentic-ai/cube/mcp otherwise select the resource through applicable Planning Analytics Workspace REST API base path.

Quick Reference – All Tools

Tool Input Arguments Output Notes
Discovery and metadata
get_available_tm1_servers None results: list[str] Retrieves a list of available TM1 server names that can be accessed.
get_tm1_cubes server: string results: list[object] Lists all cubes on a TM1 server including metadata.
list_cube_views server: string, cube: string results: list[object] Lists all views that are defined for a specified cube, including basic metadata for each view.
list_cubes_with_ai_analysis_metadata server: string, only_analyzed: boolean (optional), return_summaries: boolean (optional) results: list[object] Lists cubes on a TM1 server along with metadata that is related to AI-assisted analysis.
get_cube_dimensions_with_metadata server: string, cube: string results: list[object] Lists the dimensions of a TM1 cube along with detailed metadata for each dimension.
Dimensions and members
get_cube_dimensions server: string, cube: string results: list[string] Retrieves the list of dimension names that define a TM1 cube's structure.
get_cube_sample_members server: string, cube: string results: list[object] Retrieves a list of sample members from each dimension of a specified TM1 cube.
lookup_potential_members server: string, cube: string, search_term: string, dimension: string (optional) results: list[object] Retrieves dimension members that might correspond to a specified search term within a TM1 cube.
Candidate selection and guidance
get_cubes_that_may_answer_query server: string, query: string results: list[object] Retrieves a list of cubes from a TM1 server that might contain data that is relevant to a natural language query.
Data retrieval and execution
get_mdx_for_recommended_view server: string, query: string, cube: string (optional) cube_name: string, mdx: string, alternative_cubes: list[string], state: string Converts a natural language query into a corresponding TM1 MDX query. Works only with pre-analyzed cubes.
execute_mdx_and_get_view server: string, cube: string, mdx: string, max_rows: int (optional), max_columns: int (optional) data: string, state: string Runs a TM1 MDX query and retrieves the resulting data as a formatted markdown table.
get_data_from_data_explorer server: string, query: string, cube: string (optional), max_rows: int (optional), max_columns: int (optional) cube_name: string, filters: string, data: string, explanation: string, alternative_cubes: list[string], state: string Retrieves data from a pre-analyzed TM1 cube based on a natural language query.
View lifecycle
save_mdx_view server: string, name: string, mdx or state result: string, state: string Saves a view to a cube that uses either an MDX definition or a view state, enabling reuse of analytical context in subsequent workflows.
get_saved_view server: string, cube: string, view: string data: string, state: string Retrieves a saved view from a cube and returns its data and analytical state for further exploration or analysis.
Cube lifecycle
create_tm1_cube server: string, cube_name: string, dimensions: list[string] result: string Creates a new TM1 cube with the specified dimensions on a server.
delete_tm1_cube server: string, cube_name: string result: string Deletes an existing TM1 cube from a server.
Scenarios and sandboxes
get_tm1_sandboxes server: string results: list[object] Lists all sandboxes on a TM1 server along with associated metadata.
create_tm1_sandbox server: string, name: string, source: string (optional) result: string Creates a new TM1 sandbox on a server, optionally by using an existing sandbox as a source.
publish_sandbox server: string, name: string, clean_after: boolean (optional) result: string Publishes changes from a TM1 sandbox to base data on a server.
discard_sandbox_changes server: string, name: string result: string Discards all changes in a TM1 sandbox and restores it to its original state.
delete_tm1_sandbox server: string, name: string result: string Deletes a TM1 sandbox from a server.
Annotations
get_cube_annotations server: string, cube: string, skip: int (optional), top: int (optional) results: list[object], total_count: int, skip: int, top: int Retrieves annotations that are associated with a TM1 cube, with support for pagination.
Insights and analysis
perform_impact_analysis server: string, state: string results: list[object], status_info Performs impact analysis on cube data to identify key drivers that influence values within a specified analytical context.
get_impact_analysis_summary key_drivers: list[object] result: string Generates a natural-language summary of impact analysis results to explain key drivers and their relative influence.
perform_outlier_detection server: string, state: string results: list[object], status_info Detects anomalous values in cube data to identify rows or cells that deviate significantly from expected patterns within a specified analytical context.
get_outlier_summary outliers: list[object], algorithm: string result: string Generates a natural-language summary of detected outliers to explain unusual values or patterns that are identified in cube data.
generate_exploration_analysis_report server: string, state: string, row_range_start: int, column_range_start: int, row_range_span: int, column_range_span: int, generate_summary: boolean (optional) result: string Generates a PDF analysis report for a TM1 cube view based on a provided state and selected data ranges.