createCommunication
ChatGPTIt creates a new communication in magnews. The communication will be created considering the global:inspireemail template, and will fail if not present in your account. You should also provide an ordered list of sections to add to the communication when creating it. Use only catalog item IDs, qtypes, and qcontentblock names returned by the listTemplateSections tool. The names of the blocks are case-sensitive. Parameter values must be passed as "[qtype]": "value" entries inside params. For parameters without qcontentblock, set values directly in params. For block-specific parameters, use blocks with the exact qcontentblock name and specify the corresponding parameters within it. For sections with blocks, section-level params are used as default values for all blocks, unless overridden at block level. You don't need to specify a value for every available parameter. You can check the default values from the list of available sections for the specific template. The tool returns the ID of the newly created communication if the operation succeeds, or an error if the communication cannot be created.
createDynamicAudience
ChatGPTIt creates a new dynamic audience in magnews. The query must always follow this structure: SELECT * FROM CONTACTS WHERE ... . The fields used in the query must be fields returned by the getDatabaseFields tool. After the WHERE clause, you may define the filtering conditions. These conditions must use only field names, operators, and values. Do not use JOIN clauses or any other complex SQL operators. When a filter includes a date value, the date must be written using the following format: isodatetime('YYYY-MM-DD HH:mm:ss') This format is required so that the date can be parsed correctly. The tool returns the ID of the newly created audience if the operation succeeds, or an error if the audience cannot be created.
findAudiences
ChatGPTSearch for audiences based on specified criteria, and retrieve their basic information. This tool allows to filter one or more audiences by applying a set of clauses combined using a global boolean mode (AND / OR). The result contains informational data only (no statistics). Query semantics - Filters are expressed as a list of clauses. - All clauses are combined using a single mode: - MATCH_ALL: every clause must match (logical AND) - MATCH_ANY: at least one clause must match (logical OR) - Mixing AND and OR within the same query is not supported. - The SQL_LIKE operator behaves exactly like the SQL LIKE operator, so the '%' signs must be included in the value where appropriate. Limits - A maximum of 100 audiences can be returned. - If count is not provided, the default value (100) is used. - Requests exceeding the limit are automatically capped. Output Returns a csv, ordered by 'lastmodificationdate' (most recent first), with the following headers: "iddatabase","idaudience","audience_name","is_static","creationdate","lastmodificationdate" If an error occurs, an error string is returned instead.
findCommunications
ChatGPTSearch for communications based on specified criteria, and retrieve their full attributes. This tool allows to filter one or more communications by applying a set of clauses combined using a global boolean mode (AND / OR). The result contains informational data only (no statistics or performance metrics). Query semantics - Filters are expressed as a list of clauses. - All clauses are combined using a single mode: - MATCH_ALL: every clause must match (logical AND) - MATCH_ANY: at least one clause must match (logical OR) - Mixing AND and OR within the same query is not supported. - The SQL_LIKE operator behaves exactly like the SQL LIKE operator, so the '%' signs must be included in the value where appropriate. Limits - A maximum of 100 communications can be returned. - If count is not provided, the default value (100) is used. - Requests exceeding the limit are automatically capped. Output Returns a csv, ordered by 'sentdate' (most recent first), with the following headers: "idcommunication","name","subject","creationdate","sentdate","idjourney","idworkflow","channel","fromname","journeyname" If an error occurs, an error string is returned instead. All dates must be specified in yyyy-MM-dd HH:mm:ss format.
findWorkflows
ChatGPTSearch for workflows according to specific criteria and retrieve their full attributes. This tool allows filtering one or more workflows by applying a set of clauses combined using a global boolean mode (AND / OR). The result contains informational data along with basic sessions information. Query semantics - Filters are expressed as a list of clauses. - All clauses are combined using a single mode: - MATCH_ALL: every clause must match (logical AND) - MATCH_ANY: at least one clause must match (logical OR) - Mixing AND and OR within the same query is not supported. - The SQL_LIKE operator behaves exactly like the SQL LIKE operator, so the '%' signs must be included in the value where appropriate. Limits - A maximum of 50 workflows can be returned. - If count is not provided, the default value (50) is used. - Requests exceeding the limit are automatically capped. Output Returns a csv, ordered by 'lastpublicationdate' (most recent first) with the following headers: "idworkflow","name","description","idjourney","journeyname","creationdate","lastmodificationdate","lastpublicationdate","totalsessions","totalactivesessions","firstsessiondate","lastsessiondate","status" Deleted workflows are not included. If an error occurs, an error string is returned instead.
getAudienceFilters
ChatGPTRetrieve a SQL-like filter definitions for more than one dynamic audience for which you already know the IDs. Use this tool when you need to understand how the requested audiences are composed, including the current audience filters and the filters inherited from parent audiences. Returns a JSON object containing a list of dictionaries, one for each requested audience, with the following information: - idAudience - audienceFilter: description of the filter directly applied to the audience - parentAudiences: list of parent audiences, each containing: - idParent - parentName - filter: description of the filter inherited from that parent audience
getAudiencesInsights
ChatGPTRetrieve data and statistics/insight about more than one audiences for which you already know the IDs. Returns a JSON object containing a list of dictionaries with the following information about the requested audiences: idAudience, audienceName, audienceType, idDatabase, totalContacts, businessFields (label, average, ...), inRecency, engagementScore, conversions System audiences are redacted unless their system label is subscribed. A redacted row contains only idAudience, audienceName and a data field explaining that the audience is not visible to the user. All dates must be specified in yyyy-MM-dd HH:mm:ss format.
getCommunicationContent
ChatGPTRetrieves the name, subject, preheader, from name, from email, text content, and a screenshot of a communication. This functionality is available only for communications sent through the email channel.
getCommunicationInsights
ChatGPTRetrieve data and statistics/insight (clicks, opens, etc) about more than one communication for which you already know the IDs. Each communication can be on email or sms channel, or both. Returns a JSON object containing a list of dictionaries with information about the requested communications, along with aggregated data for all returned communications in a separate field. The response also includes statistics on communication links and conversion goals.
getContactsCount
ChatGPTCount contacts in a database or audience, optionally filtered, and return counts grouped by subscription status. Pass either iddatabase or idaudience to count contacts in that scope. These parameters are mutually exclusive, provide at most one. You may also provide filter: With `iddatabase` or `idaudience`, it further restricts contacts within that scope. Without either, it is applied across all accessible databases. filter must contain only a SQL WHERE condition, not a complete query (e.g. "sex = 'F'"). Do not include either an iddatabase or idaudience condition, use the parameters for that. Use only contact fields returned by getDatabaseFields. The result contains four counts: subscribed, unconfirmed, suspended, and unsubscribed. All four keys are always present, using 0 when no contacts match. Sum them to get the total number of contacts. On failure, the tool returns an object containing only an error key.
getDatabaseFields
ChatGPTRetrieves basic information about a database (contact database / address book), along with the list of its contact fields, grouped hierarchically by category and section. Use iddatabase = 0 to retrieve the global (shared) database fields. Set onlyGlobal to true to restrict the response to global fields only. Set onlyBusiness to true to restrict the response to business fields only. Returns a JSON object with the following top-level keys: - idDatabase: the database id - databaseName: the database name - primaryKeyField: the name of the primary-key field - fields: map of fieldName -> field descriptor (label, fieldType, fieldSubType, global, system, business, category, section, values for enum-like fields, etc.) If an error occurs, an object with a single error key is returned.
getTransactionalMessagesInsights
ChatGPTRetrieve data and statistics/insight (clicks, opens, etc) about more than one transactional message for which you already know the IDs. Each transactional message can be on email or sms channel, or both. Returns a JSON object containing a list of dictionaries with information about the requested communications, along with aggregated data for all returned communications in a separate field. The response also includes statistics on communication links and conversion goals. All dates must be specified in yyyy-MM-dd HH:mm:ss format.
getWorkflowsInsights
ChatGPTRetrieve data and statistics/insight about more than one workflows for which you already know the IDs. Returns a JSON object containing a list of dictionaries with the following information about the requested workflows: idworkflow, fromDate, toDate, totalSessions, errorSessions, sessionExits (label, exitType, count) All dates must be specified in yyyy-MM-dd HH:mm:ss format.
listDatabases
ChatGPTRetrieves the list of databases (contact databases / address books) available in the account. Output Returns a csv with the following headers: "iddatabase","name","description","primarykeyfield" If an error occurs, an error string is returned instead.
listJourneys
ChatGPTRetrieves useful data about all the journeys in the account.
listTemplateSections
ChatGPTRetrieves the list of available sections for the global:inspireemail communication template (currently the only available template). For each section, the tool provides: the section ID the section name and description * the parameters available for populating the section This information can be used to populate sections when creating or updating a communication.
updateCommunicationAttribute
ChatGPTUpdates the name, subject, or preheader of an email communication. If the communication is in draft, the name, subject, and preheader can be updated. If the communication has already been sent, only the name can be modified.
createCommunication
ClaudecreateDynamicAudience
ClaudefindAudiences
ClaudefindCommunications
ClaudefindWorkflows
ClaudegetAudienceFilters
ClaudegetAudiencesInsights
ClaudegetCommunicationContent
ClaudegetCommunicationInsights
ClaudegetContactsCount
ClaudegetDatabaseFields
ClaudegetTransactionalMessagesInsights
ClaudegetWorkflowsInsights
ClaudelistDatabases
ClaudelistJourneys
ClaudelistTemplateSections
ClaudeupdateCommunicationAttribute
Claude
