Dataiku DSS MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DSS_HOST | Yes | The URL of your Dataiku DSS instance (e.g., https://your-dss-instance.com:10000) | |
| DSS_API_KEY | Yes | Your Dataiku DSS API key | |
| DSS_INSECURE_TLS | No | Set to 'true' if using self-signed certificates (default is false) | false |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| create_recipeB | Create a new recipe in a Dataiku project |
| update_recipeC | Update an existing recipe |
| delete_recipeB | Delete a recipe from a project |
| run_recipeC | Run a recipe to build its outputs |
| create_datasetC | Create a new dataset in a project |
| update_datasetC | Update dataset settings |
| delete_datasetC | Delete a dataset from a project |
| build_datasetC | Build a dataset |
| inspect_dataset_schemaC | Get dataset schema information |
| check_dataset_metricsB | Get latest dataset metrics |
| create_scenarioB | Create a new scenario in a project |
| update_scenarioC | Update scenario settings |
| delete_scenarioB | Delete a scenario from a project |
| run_scenarioC | Run a scenario manually |
| get_scenario_logsB | Get detailed run logs and error messages for failed scenarios |
| get_recipe_codeB | Extract actual Python/SQL code from recipes |
| get_project_flowB | Get complete data flow/pipeline structure |
| get_dataset_sampleC | Get sample data from datasets |
| get_recent_runsB | Get recent run history across all scenarios/recipes |
| list_projectsB | List all available Dataiku projects |
| list_datasetsA | List all datasets in a project |
| get_dataset_infoC | Get detailed information about a dataset |
| clear_datasetC | Clear data from a dataset |
| list_recipesC | List all recipes in a project |
| get_recipe_infoB | Get detailed information about a recipe |
| validate_recipe_syntaxB | Validate Python/SQL syntax of a recipe |
| test_recipe_dry_runB | Test recipe logic without actual execution |
| list_scenariosC | List all scenarios in a project |
| get_scenario_infoC | Get detailed information about a scenario |
| add_scenario_triggerB | Add a trigger to a scenario |
| remove_scenario_triggerB | Remove a trigger from a scenario |
| get_scenario_run_historyB | Get run history for a scenario |
| get_scenario_stepsB | Get step configuration including Python code |
| clone_scenarioB | Clone an existing scenario with modifications |
| search_project_objectsB | Search for datasets, recipes, scenarios by name/pattern |
| get_code_environmentsB | List available Python/R environments |
| get_project_variablesC | Get project-level variables and configuration |
| get_connectionsC | List available data connections |
| get_job_detailsC | Get detailed job execution information |
| cancel_running_jobsC | Cancel running jobs/scenarios |
| batch_update_objectsC | Update multiple objects with similar changes |
| get_project_flowB | Get complete data flow/pipeline structure |
| export_project_configC | Export project configuration as JSON/YAML |
| duplicate_project_structureC | Copy project structure to new project |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 43 tools
Most tools have distinct purposes, e.g., get_recipe_info vs get_recipe_code clearly separate; overlapping functions like get_job_details and get_recent_runs are scoped differently. A few pairs like search_project_objects and individual list functions could cause confusion, but descriptions generally clarify.
Majority follow verb_noun pattern (e.g., create_dataset, delete_recipe). Some deviations exist (check_dataset_metrics instead of get_, batch_update_objects as a prefix), but overall consistency is high with no mixed conventions.
43 tools is high, but the Dataiku platform is complex, covering datasets, recipes, scenarios, projects, and system resources. While some tools could be consolidated (e.g., multiple scenario getters), the count is borderline but reasonable for the domain.
Covers CRUD for main objects, plus build, run, test, validate, search, batch operations, and project-level exports. Missing delete_project and user management, but core workflows are well-supported with no dead ends.