Skip to content
faros Docs
Open console

Enter a topic to search documentation.

    Troubleshoot App Studio

    Identify the failing layer before retrying a task.

    Check workspace and access

    Confirm the selected organization/workspace and that the provider is enabled there. Try a read-only operation using the same identity as the failing action. A successful administrator action does not prove another identity has access.

    Diagnose by symptom

    SymptomCheckResolution
    Project preparation cannot finishRead the missing model, repository, or template message.Fix that workspace prerequisite; return to Prepare before creating another project.
    Preview does not openCheck the bound development Instance conditions and whether the viewer is a workspace member.Resolve runtime readiness first; then correct preview access separately from production invitations.
    No Verified release / Deploy control disabledCompare the intended commit with the Publishing pipeline and open View build if present.Fix the failed component build or missing image. Use Check again after new build evidence is available.
    Production is running but a recipient is deniedCheck Manage access using the recipient identity, not an administrator session.Correct the invitation or intended visibility; do not make the app public to diagnose a private invitation.
    A removed invite still opens the appCheck workspace membership and whether a production access session already exists.Workspace members retain broad access; an existing gate session can last up to 15 minutes after grant revocation.

    Gather evidence from the CLI

    Use the failing user’s workspace context :

    Terminal
    kubectl api-resources --api-group=ai.faros.sh
kubectl auth can-i get projects.ai.faros.sh
kubectl get projects.ai.faros.sh PROJECT-NAME -o yaml
kubectl api-resources --api-group=infrastructure.faros.sh

    A missing API points to enablement/discovery; a denied operation points to authorization. If the Project is readable but its environment is stuck, trace its Infrastructure binding and inspect the referenced Instance’s conditions. Redact credentials, private URLs, and application data from collected output.

    Collect useful evidence

    Record the resource name, failing step, time, status conditions, and request/run ID where available. Share these with your operator, excluding bearer tokens, credentials, and private application data.

    If the provider itself is unavailable, use App Studio self-hosting for operator checks. Once resolved, repeat the verification step in the quickstart .