Describe the bug
We've noticed a newly injected environment variable APP_IDENTITY_ENDPOINT in our confidential container groups that aren't in the current allowlist.
internal_config.json's managedIdentity.environmentVariables currently lists only:
IDENTITY_API_VERSION
IDENTITY_HEADER
IDENTITY_SERVER_THUMBPRINT
(last updated 2026-04-17)
We've recently been seeing our deployments fail with the following error, despite no changes to our ARM template or CCE policy:
{"decision":"deny","reason":{"errors":["invalid env list: APP_IDENTITY_ENDPOINT"]}}
Workaround
Setting allow_environment_variable_dropping := true avoids the failure. Adding explicit env_rules for these variable names to our own containers does not help, as we suspect there are transient Azure-side containers that don't have this whitelisted.
Related command
az confcom
Errors
{"decision":"deny","reason":{"errors":["invalid env list: APP_IDENTITY_ENDPOINT"]}}
Issue script & Debug output
az deployment group create --name operator-control --resource-group $RG --parameters operator.parameters.json --template-file operator.json
{"status":"Failed","error":{"code":"DeploymentFailed","target":"/subscriptions/XXX/resourceGroups/uid-enclave-test/providers/Microsoft.Resources/deployments/operator-control","message":"At least one resource deployment operation failed. Please list deployment operations for details. Please see https://aka.ms/arm-deployment-operations for usage details.","details":[{"code":"ResourceDeploymentFailure","target":"/subscriptions/XXX/resourceGroups/uid-enclave-test/providers/Microsoft.ContainerInstance/containerGroups/uid-operator-5.70.159-r0-azure-cc-1","message":"The resource write operation failed to complete successfully, because it reached terminal provisioning state 'Failed'.","details":[{"message":"failed to create containerd task: failed to create shim task: failed to create container e23c0e58dd1b4dcf37466e9c9986693c51cffd9dd0c8d762af041058cced4be1: guest RPC failure: container creation denied due to policy: {\"decision\":\"deny\",\"reason\":{\"errors\":[\"invalid env list: APP_IDENTITY_ENDPOINT\"]},\"truncated\":[\"input\"]};The container group provisioning has failed. Refer to 'DeploymentFailedReason' event for more details."}]},{"code":"ResourceDeploymentFailure","target":"/subscriptions/XXX/resourceGroups/uid-enclave-test/providers/Microsoft.ContainerInstance/containerGroups/uid-operator-5.70.159-r0-azure-cc-0","message":"The resource write operation failed to complete successfully, because it reached terminal provisioning state 'Failed'.","details":[{"message":"failed to create containerd task: failed to create shim task: failed to create container 8bcb921f25ca799fd2b6fa34a5505265f232dec0ebd03081efe0efa05d128bcf: guest RPC failure: container creation denied due to policy: {\"decision\":\"deny\",\"reason\":{\"errors\":[\"invalid env list: APP_IDENTITY_ENDPOINT\"]},\"truncated\":[\"input\"]};The container group provisioning has failed. Refer to 'DeploymentFailedReason' event for more details."}]}]}}
Deployment artefacts available here: https://github.com/IABTechLab/uid2-operator/releases/tag/v5.70.159-r0
We also tried adding APP_IDENTITY_ENDPOINT to env_rules but still ran into the same error: https://github.com/IABTechLab/uid2-operator/releases#release-v5.70.159-r4
Expected behavior
The deployment succeeds
Environment Summary
azure-cli 2.88.0 *
core 2.88.0 *
telemetry 1.1.0
Extensions:
resource-graph 2.1.1
Dependencies:
msal 1.36.0
azure-mgmt-resource 24.0.0
Python location '/opt/homebrew/Cellar/azure-cli/2.88.0/libexec/bin/python'
Config directory '/Users/sean.wibisono/.azure'
Extensions directory '/Users/sean.wibisono/.azure/cliextensions'
Python (Darwin) 3.14.6 (main, Jun 10 2026, 10:03:53) [Clang 21.0.0 (clang-2100.0.123.102)]
Legal docs and information: aka.ms/AzureCliLegal
<><><>
also upgraded to 2.90.0 but still seeing the same error message
Additional context
No response
Describe the bug
We've noticed a newly injected environment variable
APP_IDENTITY_ENDPOINTin our confidential container groups that aren't in the current allowlist.internal_config.json'smanagedIdentity.environmentVariablescurrently lists only:(last updated 2026-04-17)
We've recently been seeing our deployments fail with the following error, despite no changes to our ARM template or CCE policy:
Workaround
Setting
allow_environment_variable_dropping := trueavoids the failure. Adding explicitenv_rulesfor these variable names to our own containers does not help, as we suspect there are transient Azure-side containers that don't have this whitelisted.Related command
az confcom
Errors
{"decision":"deny","reason":{"errors":["invalid env list: APP_IDENTITY_ENDPOINT"]}}Issue script & Debug output
Deployment artefacts available here: https://github.com/IABTechLab/uid2-operator/releases/tag/v5.70.159-r0
We also tried adding
APP_IDENTITY_ENDPOINTtoenv_rulesbut still ran into the same error: https://github.com/IABTechLab/uid2-operator/releases#release-v5.70.159-r4Expected behavior
The deployment succeeds
Environment Summary
azure-cli 2.88.0 *
core 2.88.0 *
telemetry 1.1.0
Extensions:
resource-graph 2.1.1
Dependencies:
msal 1.36.0
azure-mgmt-resource 24.0.0
Python location '/opt/homebrew/Cellar/azure-cli/2.88.0/libexec/bin/python'
Config directory '/Users/sean.wibisono/.azure'
Extensions directory '/Users/sean.wibisono/.azure/cliextensions'
Python (Darwin) 3.14.6 (main, Jun 10 2026, 10:03:53) [Clang 21.0.0 (clang-2100.0.123.102)]
Legal docs and information: aka.ms/AzureCliLegal
<><><>
also upgraded to 2.90.0 but still seeing the same error message
Additional context
No response