Amazon S3

S3 has no management-plane split: listing buckets and reading objects are ordinary requests signed with the same credentials. An access key is enough.

Ways to connect

• Paste a key ID and secret into the dialog.
• Drop the accessKeys.csv AWS hands you when you create an access key — the dialog reads it.
• Use what the AWS CLI already has. If a profile is configured, ObjectExplorer finds it without you typing anything.

AWS_PROFILE selects the profile, falling back to default.

The credential chain

The app walks one ordered list and takes the first source that answers, so a machine that is already set up needs no configuration here at all:

	Source	Read from
1	Environment	AWS_ACCESS_KEY_ID / AWS_SECRET_ACCESS_KEY / AWS_SESSION_TOKEN
2	Credentials file	~/.aws/credentials [profile]
3	Saved credential	what you typed into the dialog, ignored once it is close to expiring
4	Assume-role cache	the newest live entry in ~/.aws/cli/cache
5	credential_process	the helper named in ~/.aws/config — aws-vault, granted, saml2aws
6	Web identity token	AWS_WEB_IDENTITY_TOKEN_FILE + AWS_ROLE_ARN, traded at STS
7	Container role	AWS_CONTAINER_CREDENTIALS_FULL_URI / _RELATIVE_URI on ECS

A source that is not configured on this machine costs nothing and shows in the dialog as skipped rather than failed, so the report says which step is missing.

SSO

ObjectExplorer never runs aws sso login or aws login for you — a browser sign-in belongs in your terminal, where you can see what it asks. Run it yourself, and the session it writes is picked up on the next listing.

Next: Google Cloud Storage.