# Agent mode: Cancel a response Launch stage: GA `POST /api/2.0/genie/agents/{agent_id}/conversations/{conversation_id}/responses/{response_id}/cancel` Cancels an in-flight agent-mode response. `response_id` is the id returned in the `response.created` event from the agent-mode responses endpoint. The response stops at the next agent boundary and its terminal state is returned. API scopes: genie ## Path parameters - `agent_id` (string, uuid, required) The ID of the Genie agent (synonymous with the Genie space ID). Example: `e1ef34712a29169db030324fd0e1df5f` - `conversation_id` (string, uuid, required) The ID of the conversation containing the response. Example: `e1ef34712a29169db030324fd0e1df5f` - `response_id` (string, uuid, required) The ID of the response to cancel (the id from the `response.created` event). Example: `e1ef34712a29169db030324fd0e1df5f` ## Returns - `id` (string, uuid, optional, Deprecated, Output only) Message ID. Legacy identifier, use message_id instead Example: `e1ef34712a29169db030324fd0e1df5f` - `space_id` (string, uuid, required) Genie space ID Example: `e1ef34712a29169db030324fd0e1df5f` - `conversation_id` (string, uuid, required) Conversation ID Example: `e1ef34712a29169db030324fd0e1df5f` - `user_id` (int64, optional, Output only) ID of the user who created the message - `created_timestamp` (int64, optional, Output only) Timestamp when the message was created - `last_updated_timestamp` (int64, optional, Output only) Timestamp when the message was last updated - `status` (string, optional, Output only) Possible values: - `FETCHING_METADATA` - `FILTERING_CONTEXT` - `ASKING_AI` - `PENDING_WAREHOUSE` - `EXECUTING_QUERY` - `FAILED` - `COMPLETED` - `SUBMITTED` - `QUERY_RESULT_EXPIRED` - `CANCELLED` Example: `ASKING_AI` - `content` (string, required) User message content Example: `Biggest open opportunities` - `attachments` (array of object, optional, Output only) AI-generated response to the message - `text` (object, optional, Output only) Text Attachment if Genie responds with text This also contains the final summary when available. - `content` (string, optional) AI generated message - `id` (string, optional) - `purpose` (string, optional) Purpose of this text attachment. A completed message may contain more than one text attachment (for example a clarifying follow-up question alongside the final answer); use this field to tell them apart. `TEXT_ATTACHMENT_PURPOSE_ANSWER` marks the final answer/summary and `FOLLOW_UP_QUESTION` marks a clarifying question. Possible values: - `TEXT_ATTACHMENT_PURPOSE_UNSPECIFIED` - `FOLLOW_UP_QUESTION` - `TEXT_ATTACHMENT_PURPOSE_ANSWER` - `query` (object, optional, Output only) Query Attachment if Genie responds with a SQL query - `title` (string, optional) Name of the query - `query` (string, optional) AI generated SQL query - `description` (string, optional) Description of the query - `last_updated_timestamp` (int64, optional, Output only) Time when the user updated the query last - `parameters` (array of object, optional) - `keyword` (string, optional) - `value` (string, optional) - `sql_type` (string, optional) - `id` (string, optional, Deprecated, Output only) - `statement_id` (string, optional, Output only) Statement Execution API statement id. Use [Get status, manifest, and result first chunk](:method:statementexecution/getstatement) to get the full result data. - `query_result_metadata` (object, optional, Output only) Metadata associated with the query result. - `row_count` (int64, optional, Output only) The number of rows in the result set. - `is_truncated` (boolean, optional, Output only) Indicates whether the result set is truncated. - `thoughts` (array of object, optional, Public Preview, Output only) Insights into how Genie came to generate the SQL. - `thought_type` (string, optional, Public Preview, Output only) The category of this thought. The possible values are: * `THOUGHT_TYPE_DESCRIPTION`: A high-level description of how the question was interpreted. * `THOUGHT_TYPE_UNDERSTANDING`: How ambiguous parts of the question were resolved. * `THOUGHT_TYPE_DATA_SOURCING`: Which tables or datasets were identified as relevant. * `THOUGHT_TYPE_INSTRUCTIONS`: Which author-defined instructions were referenced. * `THOUGHT_TYPE_STEPS`: The logical steps taken to compute the answer. Possible values: - `THOUGHT_TYPE_UNSPECIFIED` - `THOUGHT_TYPE_DESCRIPTION` - `THOUGHT_TYPE_UNDERSTANDING` - `THOUGHT_TYPE_DATA_SOURCING` - `THOUGHT_TYPE_INSTRUCTIONS` - `THOUGHT_TYPE_STEPS` - `content` (string, optional, Public Preview, Output only) The md formatted content for this thought. - `suggested_questions` (object, optional, Output only) Follow-up questions suggested by Genie - `questions` (array of string, optional, Output only) The suggested follow-up questions - `viz` (object, optional, Output only) Visualization generated by Genie, if requested via `enable_visualization` - `title` (string, optional, Output only) Name of the visualization - `query_attachment_id` (string, optional, Output only) The ID of the query attachment the visualization was generated from - `attachment_id` (string, optional, Output only) Attachment ID - `query_result` (object, optional, Deprecated, Output only) The result of SQL query if the message includes a query attachment. Deprecated. Use `query_result_metadata` in `GenieQueryAttachment` instead. - `statement_id` (string, optional) Statement Execution API statement id. Use [Get status, manifest, and result first chunk](:method:statementexecution/getstatement) to get the full result data. - `row_count` (int64, optional) Row count of the result - `is_truncated` (boolean, optional) If result is truncated - `statement_id_signature` (string, optional) JWT corresponding to the statement contained in this result - `error` (object, optional, Output only) Error message if Genie failed to respond to the message - `error` (string, optional) - `type` (string, optional) Possible values: - `TYPE_UNSPECIFIED` - `UNEXPECTED_REPLY_PROCESS_EXCEPTION` - `GENERIC_CHAT_COMPLETION_EXCEPTION` - `CONTEXT_EXCEEDED_EXCEPTION` - `DEPLOYMENT_NOT_FOUND_EXCEPTION` - `FUNCTIONS_NOT_AVAILABLE_EXCEPTION` - `INVALID_COMPLETION_REQUEST_EXCEPTION` - `CONTENT_FILTER_EXCEPTION` - `FUNCTION_ARGUMENTS_INVALID_JSON_EXCEPTION` - `RETRYABLE_PROCESSING_EXCEPTION` - `INVALID_FUNCTION_CALL_EXCEPTION` - `LOCAL_CONTEXT_EXCEEDED_EXCEPTION` - `CHAT_COMPLETION_NETWORK_EXCEPTION` - `INVALID_CHAT_COMPLETION_JSON_EXCEPTION` - `GENERIC_CHAT_COMPLETION_SERVICE_EXCEPTION` - `WAREHOUSE_ACCESS_MISSING_EXCEPTION` - `WAREHOUSE_NOT_FOUND_EXCEPTION` - `NO_TABLES_TO_QUERY_EXCEPTION` - `SQL_EXECUTION_EXCEPTION` - `REPLY_PROCESS_TIMEOUT_EXCEPTION` - `COULD_NOT_GET_UC_SCHEMA_EXCEPTION` - `INVALID_TABLE_IDENTIFIER_EXCEPTION` - `TOO_MANY_TABLES_EXCEPTION` - `FUNCTION_ARGUMENTS_INVALID_EXCEPTION` - `GENERIC_SQL_EXEC_API_CALL_EXCEPTION` - `CHAT_COMPLETION_CLIENT_EXCEPTION` - `CHAT_COMPLETION_CLIENT_TIMEOUT_EXCEPTION` - `UNKNOWN_AI_MODEL` - `TABLES_MISSING_EXCEPTION` - `MESSAGE_DELETED_WHILE_EXECUTING_EXCEPTION` - `MESSAGE_UPDATED_WHILE_EXECUTING_EXCEPTION` - `BLOCK_MULTIPLE_EXECUTIONS_EXCEPTION` - `INVALID_CERTIFIED_ANSWER_IDENTIFIER_EXCEPTION` - `TOO_MANY_CERTIFIED_ANSWERS_EXCEPTION` - `RATE_LIMIT_EXCEEDED_GENERIC_EXCEPTION` - `RATE_LIMIT_EXCEEDED_SPECIFIED_WAIT_EXCEPTION` - `FUNCTION_CALL_MISSING_PARAMETER_EXCEPTION` - `INVALID_CERTIFIED_ANSWER_FUNCTION_EXCEPTION` - `ILLEGAL_PARAMETER_DEFINITION_EXCEPTION` - `NO_QUERY_TO_VISUALIZE_EXCEPTION` - `NO_DEPLOYMENTS_AVAILABLE_TO_WORKSPACE` - `STOP_PROCESS_DUE_TO_AUTO_REGENERATE` - `FUNCTION_ARGUMENTS_INVALID_TYPE_EXCEPTION` - `MESSAGE_CANCELLED_WHILE_EXECUTING_EXCEPTION` - `COULD_NOT_GET_MODEL_DEPLOYMENTS_EXCEPTION` - `GENERATED_SQL_QUERY_TOO_LONG_EXCEPTION` - `MISSING_SQL_QUERY_EXCEPTION` - `DESCRIBE_QUERY_UNEXPECTED_FAILURE` - `DESCRIBE_QUERY_TIMEOUT` - `DESCRIBE_QUERY_INVALID_SQL_ERROR` - `INVALID_SQL_UNKNOWN_TABLE_EXCEPTION` - `INVALID_SQL_MULTIPLE_STATEMENTS_EXCEPTION` - `INVALID_SQL_MULTIPLE_DATASET_REFERENCES_EXCEPTION` - `MESSAGE_ATTACHMENT_TOO_LONG_ERROR` - `INTERNAL_CATALOG_PATH_OVERLAP_EXCEPTION` - `INTERNAL_CATALOG_MISSING_UC_PATH_EXCEPTION` - `EXCEEDED_MAX_TOKEN_LENGTH_EXCEPTION` - `INTERNAL_CATALOG_ASSET_CREATION_ONGOING_EXCEPTION` - `INTERNAL_CATALOG_ASSET_CREATION_FAILED_EXCEPTION` - `INTERNAL_CATALOG_ASSET_CREATION_UNSUPPORTED_EXCEPTION` - `UNSUPPORTED_CONVERSATION_TYPE_EXCEPTION` - `COULD_NOT_GET_DASHBOARD_SCHEMA_EXCEPTION` - `message_id` (string, uuid, required) Message ID Example: `e1ef34712a29169db030324fd0e1df5f` - `feedback` (object, optional, Output only) User feedback for the message if provided - `rating` (string, optional) The feedback rating Possible values: - `GENIE_FEEDBACK_RATING_UNSPECIFIED` - `POSITIVE` - `NEGATIVE` - `NONE` - `comment` (string, optional, Public Preview) Optional feedback comment text ## Response ```json { "id": "e1ef34712a29169db030324fd0e1df5f", "space_id": "e1ef34712a29169db030324fd0e1df5f", "conversation_id": "e1ef34712a29169db030324fd0e1df5f", "user_id": 0, "created_timestamp": 0, "last_updated_timestamp": 0, "status": "ASKING_AI", "content": "Biggest open opportunities", "attachments": [ { "text": {}, "query": {}, "suggested_questions": {}, "viz": {}, "attachment_id": "string" } ], "query_result": { "statement_id": "string", "row_count": 0, "is_truncated": true, "statement_id_signature": "string" }, "error": { "error": "string", "type": "UNEXPECTED_REPLY_PROCESS_EXCEPTION" }, "message_id": "e1ef34712a29169db030324fd0e1df5f", "feedback": { "rating": "POSITIVE", "comment": "string" } } ```