Split an AWS ARN into its six fields or assemble one, with an explanation for every field that is legitimately empty. Nothing is contacted.
Waiting for an ARN
The sixth field keeps its own colons, so a log group ARN reads correctly.
The sixth field may contain its own colons, which is why arn:aws:logs:eu-west-1:111122223333:log-group:/aws/lambda/fn:* has a resource of log-group:/aws/lambda/fn:* rather than four separate pieces.
awsThe standard commercial partition — every region except China and GovCloud.
aws-cnThe China partition, operated separately. Accounts, IAM and ARNs do not cross into it.
aws-us-govAWS GovCloud, isolated for US government workloads.
aws-isoAn air-gapped partition for classified workloads.
aws-iso-bA second air-gapped partition.
The fields appear here
Six fields, each with what it is for — and an explanation for the ones that are legitimately empty.
An empty region or account is usually correct rather than a truncation: an S3 ARN has neither, because a bucket name is globally unique; an IAM ARN has an account but no region, because IAM is global. Each blank field says why it is blank. Nothing here contacts AWS — no credentials, no API call — so an ARN that reads perfectly may name a resource that does not exist, and the service and region lists are maintained static data rather than live.
AWS ARN Inspector runs entirely on your device — no AWS account is contacted and no credentials are needed or accepted.
You’re on the Free plan. Compare plans
The split is limited to six fields, which is the whole trick. An ARN is `arn:partition:service:region:account:resource`, and the sixth field can itself contain colons — `arn:aws:logs:eu-west-1:111122223333:log-group:/aws/lambda/fn:*` has a resource of `log-group:/aws/lambda/fn:*`. Splitting on every colon gives you eight meaningless pieces; splitting on the first five gives you the right answer.
An empty field is usually correct rather than a truncation. An S3 ARN has no region and no account, because a bucket name is globally unique and neither is needed to identify it. An IAM ARN has an account but no region, because IAM is global. People routinely read those blanks as a broken ARN and go looking for the missing part, so each one carries a note explaining why it is empty.
The partition matters more than it looks. `aws-cn` and `aws-us-gov` are separate partitions: accounts, IAM principals and ARNs do not cross between them, so a policy written with the wrong partition silently matches nothing at all.
Building an ARN exposes the separator as a choice, with the honest caveat that it is not one — whether a service joins the resource type and name with a slash or a colon is a property of that service. S3 and IAM use a slash; CloudWatch Logs and SNS use a colon. The wrong one produces an ARN that is syntactically fine and matches nothing.
Nothing is contacted. No AWS account, no credentials, no API call — so an ARN that reads perfectly may name a resource that does not exist.
Because that is correct. A bucket name is globally unique, so neither a region nor an account is needed to identify it — `arn:aws:s3:::my-bucket` is complete, not truncated. The tool notes which services legitimately leave those blank.
Because the sixth field may contain colons of its own. A CloudWatch log group ARN ends `log-group:/aws/lambda/fn:*`, which is one resource and not three fields. The split has to stop after the fifth colon, which is what this does.
Yes, and it is a quiet failure when it is wrong. `aws`, `aws-cn` and `aws-us-gov` are separate partitions whose accounts and IAM principals do not cross. A policy referencing the wrong partition is valid JSON that matches nothing.