Actions, resources, and condition keys for AWS DeepRacer

AWS DeepRacer (service prefix: deepracer) provides the following service-specific resources, actions, and condition context keys for use in IAM permission policies.

References:

• Learn how to configure this service.

• View a list of the API operations available for this service.

• Learn how to secure this service and its resources by using IAM permission policies.

Actions defined by AWS DeepRacer

You can specify the following actions in the Action element of an IAM policy statement. Use policies to grant permissions to perform an operation in AWS. When you use an action in a policy, you usually allow or deny access to the API operation or CLI command with the same name. However, in some cases, a single action controls access to more than one operation. Alternatively, some operations require several different actions.

The Resource types column of the Actions table indicates whether each action supports resource-level permissions. If there is no value for this column, you must specify all resources ("*") to which the policy applies in the Resource element of your policy statement. If the column includes a resource type, then you can specify an ARN of that type in a statement with that action. If the action has one or more required resources, the caller must have permission to use the action with those resources. Required resources are indicated in the table with an asterisk (*). If you limit resource access with the Resource element in an IAM policy, you must include an ARN or pattern for each required resource type. Some actions support multiple resource types. If the resource type is optional (not indicated as required), then you can choose to use one of the optional resource types.

The Condition keys column of the Actions table includes keys that you can specify in a policy statement's Condition element. For more information on the condition keys that are associated with resources for the service, see the Condition keys column of the Resource types table.

Note

Resource condition keys are listed in the Resource types table. You can find a link to the resource type that applies to an action in the Resource types (*required) column of the Actions table. The resource type in the Resource types table includes the Condition keys column, which are the resource condition keys that apply to an action in the Actions table.

For details about the columns in the following table, see Actions table.

Actions	Description	Access level	Resource types (*required)	Condition keys	Dependent actions
AddLeaderboardAccessPermission [permission only]	Grants permission to add access for a private leaderboard	Write	leaderboard*
				deepracer:UserToken deepracer:MultiUser
AdminGetAccountConfig [permission only]	Grants permission to get current admin multiuser configuration for this account	Read
AdminListAssociatedResources [permission only]	Grants permission to list all deepracer users with their associated resources created under this account	Read
AdminListAssociatedUsers [permission only]	Grants permission to list user data for all users associated with this account	Read
AdminManageUser [permission only]	Grants permission to manage a user associated with this account	Write
AdminSetAccountConfig [permission only]	Grants permission to set configuration options for this account	Write
CloneReinforcementLearningModel [permission only]	Grants permission to clone an existing DeepRacer model	Write	reinforcement_learning_model*
			track*
				aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
CreateCar [permission only]	Grants permission to create a DeepRacer car in your garage	Write		aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
CreateLeaderboard [permission only]	Grants permission to create a leaderboard	Write		aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
CreateLeaderboardAccessToken [permission only]	Grants permission to create an access token for a private leaderboard	Write	leaderboard*
				deepracer:UserToken deepracer:MultiUser
CreateLeaderboardSubmission [permission only]	Grants permission to submit a DeepRacer model to be evaluated for leaderboards	Write	leaderboard*
			reinforcement_learning_model*
				aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
CreateReinforcementLearningModel [permission only]	Grants permission to create ra einforcement learning model for DeepRacer	Write	track*
				aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
DeleteLeaderboard [permission only]	Grants permission to delete a leaderboard	Write	leaderboard*
				deepracer:UserToken deepracer:MultiUser
DeleteModel [permission only]	Grants permission to delete a DeepRacer model	Write	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
EditLeaderboard [permission only]	Grants permission to edit a leaderboard	Write	leaderboard*
				deepracer:UserToken deepracer:MultiUser
GetAccountConfig [permission only]	Grants permission to get current multiuser configuration for this account	Read		deepracer:UserToken deepracer:MultiUser
GetAlias [permission only]	Grants permission to retrieve the user's alias for submitting a DeepRacer model to leaderboards	Read		deepracer:UserToken deepracer:MultiUser
GetAssetUrl [permission only]	Grants permission to download artifacts for an existing DeepRacer model	Read	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
GetCar [permission only]	Grants permission to retrieve a specific DeepRacer car from your garage	Read	car*
				deepracer:UserToken deepracer:MultiUser
GetCars [permission only]	Grants permission to view all the DeepRacer cars in your garage	Read		deepracer:UserToken deepracer:MultiUser
GetEvaluation [permission only]	Grants permission to retrieve information about an existing DeepRacer model's evaluation jobs	Read	evaluation_job*
				deepracer:UserToken deepracer:MultiUser
GetLatestUserSubmission [permission only]	Grants permission to retrieve information about how the latest submitted DeepRacer model for a user performed on a leaderboard	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
GetLeaderboard [permission only]	Grants permission to retrieve information about leaderboards	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
GetModel [permission only]	Grants permission to retrieve information about an existing DeepRacer model	Read	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
GetPrivateLeaderboard [permission only]	Grants permission to retrieve information about private leaderboards	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
GetRankedUserSubmission [permission only]	Grants permission to retrieve information about the performance of a user's DeepRacer model that got placed on a leaderboard	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
GetTrack [permission only]	Grants permission to retrieve information about DeepRacer tracks	Read	track*
GetTrainingJob [permission only]	Grants permission to retrieve information about an existing DeepRacer model's training job	Read	training_job*
				deepracer:UserToken deepracer:MultiUser
ImportModel [permission only]	Grants permission to import a reinforcement learning model for DeepRacer	Write		deepracer:UserToken deepracer:MultiUser
ListEvaluations [permission only]	Grants permission to list a DeepRacer model's evaluation jobs	Read	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
ListLeaderboardEvaluations [permission only]	Grants permission to list all the user's leaderboard evaluation jobs for a leaderboard	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
ListLeaderboardSubmissions [permission only]	Grants permission to list all the DeepRacer model submissions of a user on a leaderboard	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
ListLeaderboards [permission only]	Grants permission to list all the available leaderboards	Read		deepracer:UserToken deepracer:MultiUser
ListModels [permission only]	Grants permission to list all existing DeepRacer models	Read		deepracer:UserToken deepracer:MultiUser
ListPrivateLeaderboardParticipants [permission only]	Grants permission to retrieve participant information about private leaderboards	Read	leaderboard*
				deepracer:UserToken deepracer:MultiUser
ListPrivateLeaderboards [permission only]	Grants permission to list all the available private leaderboards	Read		deepracer:UserToken deepracer:MultiUser
ListSubscribedPrivateLeaderboards [permission only]	Grants permission to list all the subscribed private leaderboards	Read		deepracer:UserToken deepracer:MultiUser
ListTagsForResource	Grants permission to lists tag for a resource	Read	car
			evaluation_job
			leaderboard
			leaderboard_evaluation_job
			reinforcement_learning_model
			training_job
				aws:ResourceTag/${TagKey} deepracer:UserToken deepracer:MultiUser
ListTracks [permission only]	Grants permission to list all DeepRacer tracks	Read
ListTrainingJobs [permission only]	Grants permission to list a DeepRacer model's training jobs	Read	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
MigrateModels [permission only]	Grants permission to migrate previous reinforcement learning models for DeepRacer	Write
PerformLeaderboardOperation [permission only]	Grants permission to performs the leaderboard operation mentioned in the operation attribute	Write	leaderboard
				deepracer:UserToken deepracer:MultiUser
RemoveLeaderboardAccessPermission [permission only]	Grants permission to remove access for a private leaderboard	Write	leaderboard*
				deepracer:UserToken deepracer:MultiUser
SetAlias [permission only]	Grants permission to set the user's alias for submitting a DeepRacer model to leaderboards	Write		deepracer:UserToken deepracer:MultiUser
StartEvaluation [permission only]	Grants permission to evaluate a DeepRacer model in a simulated environment	Write	reinforcement_learning_model*
			track*
				aws:RequestTag/${TagKey} aws:TagKeys deepracer:UserToken deepracer:MultiUser
StopEvaluation [permission only]	Grants permission to stop DeepRacer model evaluations	Write	evaluation_job*
				deepracer:UserToken deepracer:MultiUser
StopTrainingReinforcementLearningModel [permission only]	Grants permission to stop training a DeepRacer model	Write	reinforcement_learning_model*
				deepracer:UserToken deepracer:MultiUser
TagResource	Grants permission to tag a resource	Tagging	car
			evaluation_job
			leaderboard
			leaderboard_evaluation_job
			reinforcement_learning_model
			training_job
				aws:TagKeys aws:RequestTag/${TagKey} aws:ResourceTag/${TagKey} deepracer:UserToken deepracer:MultiUser
TestRewardFunction [permission only]	Grants permission to test reward functions for correctness	Write
UntagResource	Grants permission to untag a resource	Tagging	car
			evaluation_job
			leaderboard
			leaderboard_evaluation_job
			reinforcement_learning_model
			training_job
				aws:TagKeys deepracer:UserToken deepracer:MultiUser
UpdateCar [permission only]	Grants permission to update a DeepRacer car in your garage	Write	car*
				deepracer:UserToken deepracer:MultiUser

Resource types defined by AWS DeepRacer

The following resource types are defined by this service and can be used in the Resource element of IAM permission policy statements. Each action in the Actions table identifies the resource types that can be specified with that action. A resource type can also define which condition keys you can include in a policy. These keys are displayed in the last column of the Resource types table. For details about the columns in the following table, see Resource types table.

Resource types	ARN	Condition keys
car	arn:${Partition}:deepracer:${Region}:${Account}:car/${ResourceId}	aws:ResourceTag/${TagKey}
evaluation_job	arn:${Partition}:deepracer:${Region}:${Account}:evaluation_job/${ResourceId}	aws:ResourceTag/${TagKey}
leaderboard	arn:${Partition}:deepracer:${Region}::leaderboard/${ResourceId}	aws:ResourceTag/${TagKey}
leaderboard_evaluation_job	arn:${Partition}:deepracer:${Region}:${Account}:leaderboard_evaluation_job/${ResourceId}	aws:ResourceTag/${TagKey}
reinforcement_learning_model	arn:${Partition}:deepracer:${Region}:${Account}:model/reinforcement_learning/${ResourceId}	aws:ResourceTag/${TagKey}
track	arn:${Partition}:deepracer:${Region}::track/${ResourceId}
training_job	arn:${Partition}:deepracer:${Region}:${Account}:training_job/${ResourceId}	aws:ResourceTag/${TagKey}

Condition keys for AWS DeepRacer

AWS DeepRacer defines the following condition keys that can be used in the Condition element of an IAM policy. You can use these keys to further refine the conditions under which the policy statement applies. For details about the columns in the following table, see Condition keys table.

To view the global condition keys that are available to all services, see Available global condition keys.

Condition keys	Description	Type
aws:RequestTag/${TagKey}	Filters actions by tag key-value pairs in the request	String
aws:ResourceTag/${TagKey}	Filters actions by tag key-value pairs attached to the resource	String
aws:TagKeys	Filters actions by tag keys in the request	ArrayOfString
deepracer:MultiUser	Filters access by multiuser flag	Bool
deepracer:UserToken	Filters access by user token in the request	String