Every action below is available as an MCP tool and a verb in Cerebral OS — callable from any AI agent, Claude, Cursor, Windsurf, or your own runtime via the BYOA API. All executions are governed, audited, and dry-run safe.
Create Dataset
relevance_ai:create_dataset
Create a new dataset with optional schema definition.
Write
Medium risk
Create Vector Field
relevance_ai:create_vector_field
Create a vector field in a dataset by vectorizing text from an existing field.
Write
Medium risk
Delete Dataset
relevance_ai:delete_dataset
Delete a dataset permanently. This action cannot be undone.
Write
High risk
Delete Document
relevance_ai:delete_document
Delete a document from a dataset permanently. This action cannot be undone.
Write
High risk
Get Dataset
relevance_ai:get_dataset
Fetch a single dataset by ID with schema and metadata information.
Read
Low risk
Get Document
relevance_ai:get_document
Retrieve a single document by ID from a dataset.
Read
Low risk
Get Job Status
relevance_ai:get_job_status
Check the status of an asynchronous job (workflow run, vectorization, etc.).
Read
Low risk
Insert Documents
relevance_ai:insert_documents
Insert documents into a dataset. Documents will be validated against the dataset schema.
Write
Medium risk
List Datasets
relevance_ai:list_datasets
List all datasets with optional filtering and pagination.
Read
Low risk
List Workflows
relevance_ai:list_workflows
List all available workflows with optional filtering and pagination.
Read
Low risk
Run Workflow
relevance_ai:run_workflow
Execute a workflow with optional input parameters.
Write
Medium risk
Search Documents
relevance_ai:search_documents
Search documents in a dataset using text or vector similarity.
Read
Low risk
Update Document
relevance_ai:update_document
Update an existing document in a dataset with new field values.
Write
Medium risk