mallochio/openml-mcp-server

get_task_description

Get the description for a specific OpenML task by its ID. Describes ML tasks like classification or regression, including inputs and evaluation criteria.

Args: task_id: The integer ID of the task.

list_tasks

List OpenML tasks, applying filters specified as part of the path. Example filters: 'limit/10/offset/0', 'type/1/tag/study_1' (type 1 is Supervised Classification). See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/offset/0', 'type/1').

get_flow_description

Get the description for a specific OpenML flow (model/pipeline) by its ID. Includes parameters, dependencies, etc.

Args: flow_id: The integer ID of the flow.

list_flows

List OpenML flows, applying filters specified as part of the path. Example filters: 'limit/10/offset/0', 'tag/weka'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/offset/0', 'uploader/1').

check_flow_exists

Check if a flow with a specific name and external version exists on OpenML.

Args: name: The name of the flow (e.g., 'weka.J48'). version: The external version string (e.g., 'Weka_3.7.5_9117').

get_run_description

Get the description for a specific OpenML run (experiment result) by its ID. Includes task, flow, setup, evaluations, and output files.

Args: run_id: The integer ID of the run.

list_runs

List OpenML runs, applying filters specified as part of the path. Requires filters like task, flow, setup, or uploader ID. Max 10,000 results. Example filters: 'limit/10/offset/0/task/1', 'flow/67/uploader/1'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/task/28', 'flow/67'). Must include task, flow, setup, uploader, or run filter.

get_run_trace

Get the optimization trace for a specific OpenML run (if available). Shows hyperparameter settings tried during tuning and their evaluations.

Args: run_id: The integer ID of the run.

list_evaluations

List OpenML run evaluations, applying filters specified as part of the path. Requires filters like function, task, flow, setup, uploader, or run ID. Max 10,000 results. Example filters: 'limit/10/offset/0/task/1/function/predictive_accuracy'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/task/68/function/f_measure'). Must include filters.

get_setup_description

Get the description for a specific OpenML setup (hyperparameter configuration) by its ID.

Args: setup_id: The integer ID of the setup.

list_setups

List OpenML setups, applying filters specified as part of the path. Max 1,000 results. Example filters: 'limit/10/offset/0/flow/65'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/flow/65', 'setup/10,12').

get_study_description

Get the description for a specific OpenML study (collection of tasks/runs) by its ID or alias.

Args: study_id_or_alias: The integer ID or string alias of the study.

list_studies

List OpenML studies, applying filters specified as part of the path. Example filters: 'limit/10/offset/0', 'main_entity_type/task'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/offset/0', 'main_entity_type/task'). Can be empty for all studies.

list_task_types

List all task types supported by OpenML (e.g., Supervised Classification, Regression).

get_task_type_description

Get the description for a specific OpenML task type by its ID.

Args: task_type_id: The integer ID of the task type (e.g., 1 for Supervised Classification).

list_evaluation_measures

List all evaluation measures supported by OpenML (e.g., predictive_accuracy, area_under_roc_curve).

list_estimation_procedures

List all estimation procedures supported by OpenML (e.g., 10-fold Crossvalidation).

get_dataset_description

Get the description for a specific OpenML dataset by its ID. Contains all the meta-data about the dataset.

Args: dataset_id: The integer ID of the dataset.

list_datasets

List OpenML datasets, applying filters specified as part of the path. Example filters: 'limit/10/offset/0', 'status/active/tag/uci', 'number_instances/0..1000'. See OpenML API docs for detailed filter syntax.

Args: filters: The filter string (e.g., 'limit/10/offset/0', 'status/active').

get_dataset_features

Get the features (attributes/columns) description for a specific OpenML dataset.

Args: dataset_id: The integer ID of the dataset.

get_dataset_qualities

Get the calculated qualities (meta-features) for a specific OpenML dataset.

Args: dataset_id: The integer ID of the dataset.

list_data_qualities

List all available data quality measures supported by OpenML.