A new mutation lets users export multi-organization usage reports. The report is generated asynchronously and delivered by email, mirroring the single-organization Usage Stats export.

Added:

• The mutation createMultiOrgUsageReportExport triggers an asynchronous usage report export aggregated across multiple organizations. Input:
  • organizationUuids: [ID!]! — the organizations to include (1..25); only those the caller may see usage stats for are included.
  • period: PeriodFilter! — current_month, last_month, or last_3_months.
  • resourceType: MultiOrgUsageResourceType! — api (AI credits and automations follow).
  • Returns { status } (created on a successful trigger).

Notes:

• The mutation is gated by the multi_org_usage_stats feature flag; it is hidden from introspection and rejected unless the flag is enabled for the caller's organization.
• The report is emailed to the requesting user when ready — there is no polling query in this release.
• resourceType accepts ai_credit and automation as values, but only api produces a report at this time; the others are rejected until their exporters ship.

The mcpServerUsages query now reports usage across every resource that references an MCP server, so integrations can warn about all dependent behaviors before deleting a connection.

Changed:

• The query mcpServerUsages is now keyed solely on mcpServerId and returns usage across all resources that reference the server, not only the pipe passed in resource. The resource argument is still required for authorization but no longer scopes the result.

Notes:

• Use mcpServerUsages before deleting an MCP server (deleteMcpServer) to surface the behaviors (and their agents) that will lose the server's tools. Deletion is not blocked by usage — the query is informational so you can warn the user first.

The March changelog introduces significant enhancements to LLM provider management through new specialized queries and mutations, alongside the addition of web scraping capabilities to AI behaviors. Key structural updates include the implementation of deletedAt fields for better deletion tracking across various objects and a critical type change from Int to Float for AI usage and credit fields. These updates expand metadata tracking while requiring minor adjustments to existing integrations to ensure data type precision.

New:

• A new query activeLlmProvider was added to fetch details of the currently active LLM provider for a given resource.
• A new query llmProvidersByOrganization was added to retrieve LLM providers by organization.
• A new query providerDependencies was added to get information about provider dependencies.
• The AiBehaviorCapabilityType enum received a new value web_scraping.
• Fields phase, pipe, table, and tableId were added to AiBehaviorMetadata for expanded metadata tracking.
• A field dependentInterfaces was added to FieldDependency.
• Newly added fields id to both FlowAiAgent and FlowAutomation for unique identification.
• Several object types received a new field deletedAt for tracking deletions:
  • ConnectedTable
  • InternalPipe
  • Phase
  • PhaseFlow
  • Pipe
  • Table
  • Repo interface
• New mutations for managing LLM providers:
  • createLlmProvider
  • deleteLlmProvider
  • resetLlmProviderOwner
  • setActiveLlmProvider
  • updateLlmProvider
• A new mutation duplicateCondition was added to duplicate field conditions.

Breaking Changes:

• The fields AgentsDetails.usage, AiResource.usage, StatsDetails.usage, and UsageStatsAiCreditDetails.usage changed type from Int! to Float!. If you're using these fields, ensure to handle float values instead of integers.
• The field FieldDependency.dependentAiAgents has been updated to use DependentAiAgentConnection instead of AutomationConnection.
• The field FieldDependency.dependentAutomations now expects a non-null return with AutomationConnection!.

These updates introduce several new capabilities, particularly around LLM provider management, while also making sure your current interactions with fields are precise in their data types. If you have any integrations using the updated or changed fields, ensure they are properly adapted to accommodate these updates.

This month's update introduces enhancements to the GraphQL API, including a new query for the Flow view, identity propagation configuration capabilities, field archiving support, and improved notification handling for automation loops.

New:

• Query Added:

  • pipeFlow: A new query responsible for returning data to be shown on the new Flow view.

• Mutation Added:

  • createIdentityPropagationConfiguration: A new mutation for creating identity propagation configurations.

• Field Added: The archived field was introduced across all pipefy fields (e.g. AssigneeField, DateField) to indicate whether the field was archived.

• Enum Value: automationLoopDetected was added to the enum NotificationEvent. You can now use this value to handle specific notification cases related to automation loops.

This release introduces Start Form visibility controls, IPaaS preferences, and archived status fields for Phases and Reports. Additionally, we have deprecated the anyone_can_create_card field across multiple objects and removed enableExternalGuests from Repo preferences.

New:

• Fields Added:
  • archived was added to the PhaseField object type.
  • archived was added to the PipeReportFilterableFields object type.
  • startFormVisibility was added to the Pipe object type.
  • startFormVisibility was added to the PipeSharedProperties interface.

Deprecated:

• The anyone_can_create_card field is now deprecated across several areas including:
  • ConnectedTable
  • InternalPipe
  • Pipe
  • PublicPipeInterface
  • Repo
  • Table
• The enableExternalGuests field is deprecated in RepoPreference.
• The anyone_can_create_card argument in the pipes field of the following is now deprecated:
  • InternalOrganization
  • Organization
  • OrganizationInterface
• The anyone_can_create_card input field in UpdatePipeInput is now deprecated.

These permissions are deprecated in favor of form permissions configuration.

This week brings enhanced access control with custom role management and improved compliance through pipeline audit log exports.

New Mutations:

• createCustomRole - Creates a custom role with specified title, description, and reference role
• deleteCustomRole - Deletes a custom role by its ID
• saveRolePermission - Updates the permissions assigned to a role
• exportPipeAuditLogsReport - Exports pipeline audit logs for compliance and reporting purposes

New Queries:

• availableRoles - Retrieves available roles for a resource (Repo, Organization, or Interface) based on user permissions

Feel free to explore these additions to enhance your use of the API!

This week's API updates deliver a complete pipe reporting feature set that includes three new mutations (createPipeReport, updatePipeReport, deletePipeReport) and enhanced data retrieval capabilities through three new queries (pipeReports, pipeReportColumns, pipeReportFilterableFields) with its GraphQL types, inputs and payloads.

New:

Mutations

• createPipeReport: A mutation to create new pipe reports.
• deletePipeReport: A mutation to delete existing pipe reports.
• updatePipeReport: A mutation to update existing pipe reports.

Queries

• pipeReportColumns: A query to retrieve pipe report columns.
• pipeReportFilterableFields: A query to identify which fields are filterable in pipe reports.
• pipeReports: A query to fetch a list of pipe reports.

Types

• CreatePipeReportInput: Input type for creating pipe reports.
• CreatePipeReportPayload: Payload type returned when a pipe report is created.
• DeletePipeReportInput: Input type for deleting pipe reports.
• DeletePipeReportPayload: Payload type returned when a pipe report is deleted.
• PipeReportConnection: Describes the connection for pipe reports.
• PipeReportEdge: Represents an edge in a connection of pipe reports.
• PipeReportFieldOptionsGQL: Offers field options for pipe reports.
• PipeReportFilterableFields: Contains filterable fields for pipe reports.
• PipeReportFilterableGQLType: Defines filterable GraphQL types for pipe reports.
• PipeReportFilterablePhasesGQL: Specifies filterable phases for pipe reports.
• PipeReportSortField: Describes fields available for sorting pipe reports.
• PipeReportsOrder: Defines the order of pipe reports.
• Sort: Specifies sorting options.
• UpdatePipeReportInput: Input type for updating pipe reports.
• UpdatePipeReportPayload: Payload type returned when a pipe report is updated.

This release enables advanced filtering by phases and field types, flexible sorting and pagination, and comprehensive report management—giving you everything needed to build powerful, customizable reporting interfaces with complete programmatic control over your pipe data analytics.

Enjoy the new features!

This week's updates bring enhanced querying options for cards and new fields for tracking AI credits and card submission information.

New:

• Fields and Arguments:

  • aiCreditsAddonAvailable field added to:

    • Organization
    • OrganizationInterface

  • submitterEmail argument added to the createCard mutation to capture the email of public form submitters.

  • pipeRelationId argument added to the cards query for authorization purposes when listing cards through pipe relations.

  • throughConnectors argument added to the cards query to filter cards through connected fields.

These updates expand query capabilities and provide better control over card creation workflows and organization data access.

This week's update makes it easier to manage AI-related features. We've added a new generatedByAi field on the NodeFieldValueInput to mark values created by AI, and we've added ai_credits to the SubscriptionLimits enum to give you more control over AI subscriptions.

New:

• Input field generatedByAi of type Boolean: This field is now available in the NodeFieldValueInput input object type. You can use it to indicate if a field value was generated by AI.

• Enum value ai_credits: Added to the SubscriptionLimits enum, this new value might help in managing AI-related subscription capabilities.

This update adds new fields to help you manage AI features and SMTP configurations. You can now track who disabled AI features and why, check for SMTP reauthentication, and specify reference connectors when creating cards.

New:

• Field aiDisabledBy has been added to the OrganizationPreferences type. This new field can help you know who disabled AI-related features.
• Field aiDisabledReason has also been added to the OrganizationPreferences type. It provides insights into why AI-related features might be disabled.
• Field reauthRequired is now part of the SmtpCustomEmail type. It indicates whether reauthentication is required for SMTP custom emails.
• Field hasAddon has been introduced to the UsageStatsAiCreditDetails type. This field helps to identify if there's an addon associated with AI credit usage.
• Input field throughConnectors has been added to the CreateCardInput input object type. This facilitates specifying reference connectors when creating a card.

These updates enhance information transparency and connectivity within the API, helping users better manage organizational settings and usage statistics.