6.64.0 (September 9, 2026)
FEATURES:
- New Data Source:
aws_accountaccess_application(#49553) - New Data Source:
aws_accountaccess_entitlements(#49554) - New Data Source:
aws_agentregistry_registry(#49806) - New Data Source:
aws_rds_events(#49783) - New List Resource:
aws_accountaccess_entitlement(#49552) - New List Resource:
aws_agentregistry_registry(#49549) - New List Resource:
aws_bedrock_model_invocation_job(#49877) - New List Resource:
aws_dms_data_provider(#49897) - New List Resource:
aws_dms_instance_profile(#49747) - New List Resource:
aws_fis_safety_lever_state(#49841) - New List Resource:
aws_lambda_alias(#49706) - New Resource:
aws_accountaccess_entitlement(#49552) - New Resource:
aws_agentregistry_registry(#49549) - New Resource:
aws_bedrock_model_invocation_job(#49877) - New Resource:
aws_dms_data_provider(#49897) - New Resource:
aws_dms_instance_profile(#49747) - New Resource:
aws_fis_safety_lever_state(#49841)
ENHANCEMENTS:
- data-source/aws_workspaces_directory: Add
workspace_access_properties.access_endpoint_configattribute (#49849) - resource/aws_appconfig_hosted_configuration_version: Add
version_labelargument (#49881) - resource/aws_bedrockagentcore_harness: Add
model.bedrock_model_config.additional_paramsargument (#48498) - resource/aws_bedrockagentcore_harness: Add
model.bedrock_model_config.api_formatandmodel.openai_model_config.api_formatarguments (#48521) - resource/aws_bedrockagentcore_harness: Add
model.gemini_model_config.additional_paramsandmodel.openai_model_config.additional_paramsarguments (#48656) - resource/aws_bedrockagentcore_harness: Add
model.litellm_model_configconfiguration block (#48656) - resource/aws_bedrockagentcore_harness: Add
skill.aws_skills,skill.git, andskill.s3configuration blocks (#48656) - resource/aws_bedrockagentcore_harness: Change
max_tokensto Optional and Computed (#48656) - resource/aws_bedrockagentcore_harness: Change
skill.pathto Optional (#48656) - resource/aws_bedrockagentcore_harness: Change
system_promptto Required (#48656) - resource/aws_bedrockagentcore_oauth2_credential_provider: Add Resource Identity support (#49900)
- resource/aws_bedrockagentcore_oauth2_credential_provider: Add configurable resource timeouts (#49900)
- resource/aws_cloudwatch_metric_alarm: Add
warm_up_configurationconfiguration block (#49873) - resource/aws_db_instance: Add
warning_event_categoriesargument. When set, surface a warning diagnostic for each matching RDS event reported during create or update. Requires therds:DescribeEventsIAM permission (#49783) - resource/aws_ecs_capacity_provider: Add
managed_instances_provider.auto_repair_configurationconfiguration block (#49763) - resource/aws_msk_replicator: Add
kafka_cluster.client_authenticationto configuremtlsorsasl_scramauthentication to an Apache Kafka cluster (#49265) - resource/aws_msk_replicator: Add
kafka_cluster.encryption_in_transitto supply a custom root CA certificate for an Apache Kafka cluster (#49265) - resource/aws_msk_replicator: Support self-managed and on-premises Apache Kafka clusters as a replication source or target, via the
kafka_cluster.apache_kafka_clusterblock and thereplication_info_list.source_kafka_cluster_idandtarget_kafka_cluster_idarguments (#49265) - resource/aws_rds_cluster: Add
warning_event_categoriesargument. When set, surface a warning diagnostic for each matching RDS event reported during create or update. Requires therds:DescribeEventsIAM permission (#49783) - resource/aws_rds_cluster_instance: Add
warning_event_categoriesargument. When set, surface a warning diagnostic for each matching RDS event reported during create or update. Requires therds:DescribeEventsIAM permission (#49783)
BUG FIXES:
- data-source/aws_workspaces_directory: Fix
setting workspace_access_properties: Invalid address to seterrors (#49849) - resource/aws_bedrockagentcore_harness: Retry
Role validation failed for '...'. Please verify that the role exists and its trust policy allows assumption by this serviceIAM eventual consistency errors on Create. Because this error is returned while waiting for a newly-created harness to stabilize, the failed harness is deleted and creation is restarted. You may see CloudTrail events that reflect this sequence of operations (#48656) - resource/aws_db_instance: Fix error when restoring multi-az SQL Server from snapshot. (#49846)
- resource/aws_rds_global_cluster: Fix potential eventual consistency error when re-creating resource. (#49804)