Keycloak

Services authenticate with the Keycloak client they run as.

Identity proof

The token has to come from https://sso.acme.com/realms/prod, and its azp claim has to match the workload you named. We compare that claim byte for byte, with no wildcards.

azp is matched rather than sub, which on this platform does not name the workload in a form you can pin.

Every token also has to name your workspace as its audience. Ask for this audience and no other. A token requested for two audiences fails every exchange.

Console fields

The console also asks which project and environment this identity reaches and which role it gets, with the credential lifetime behind Advanced.

Trust

Built from the example answers above:

We name the identity for you, so the form never asks for one.

Console snippet

# Client → Client scopes → dedicated → Add mapper → Audience: YOUR_WORKSPACE_ID
# Then turn OFF Full scope allowed there, or remove the roles scope: its audience-resolve
# mapper otherwise adds `account` and the token arrives with two audiences.
# Ask for this audience and no other. A token requested for two audiences fails every exchange.
# $KEYCLOAK_CLIENT_SECRET is this client's own secret, already on the machine.
T=$(printf 'grant_type=client_credentials&client_id=%s&client_secret=%s' \
  "$KEYCLOAK_CLIENT_ID" "$KEYCLOAK_CLIENT_SECRET" \
  | curl -sS <ISSUER>/protocol/openid-connect/token --data-binary @- | jq -r .access_token)
C=$(T="$T" jq -n '{token:env.T}' | curl -sS "https://penv.cloud/api/v1/auth/oidc" \
  -H 'content-type: application/json' --data-binary @- | jq -r .credential)
PENV_TOKEN="$C" penv pull

The snippet finishes with penv pull, which writes a file. We recommend penv run for a process; penv pull is for a host that needs a file on disk. Connect your CI has the wrapped step.

Related

• Machine identity
• Connect a Platform in the console
• Keycloak documentation: the subject is the value most often set wrong