Usage
Run it as gluon models approve (in the repository, cargo run -p gluon-cli -- models approve).
Approve a held Model version without a shadow run, through the API: audited with the reason; needs a manage-security admin's token (`--token` or GLUON_TOKEN)
Usage: gluon models approve [OPTIONS] --reason <REASON> <MODEL>
Arguments:
<MODEL> The Model, like model:eu-gpt-5-2026-08
Options:
--reason <REASON> Why it is approved without a shadow run
--url <URL> The API (GLUON_URL, else http://127.0.0.1:8420)
--token <TOKEN> A manage-security admin's token (GLUON_TOKEN)
-h, --help Print help
What it reads and writes
It calls POST /api/admin/platform/model-versions/approve on the API given with --url (else GLUON_URL, default http://127.0.0.1:8420) with the token given with --token (else GLUON_TOKEN), as an admin with manage-security. The reason (1 to 1,000 characters) goes into the audit record admin.model-version.approve, and routing uses that version at once. It prints the server's answer as JSON: {"model","version","approvedAt","approvedBy","approvedByName"}. When the server refuses, it prints the refusal: 403 without manage-security, 404 when the Model is not in the approved matrix, 409 not-held when that version is approved already, 422 when the reason is empty or too long.
Before you run it
The override is for demo stacks and test environments that cannot spend a shadow run's tokens (Waseem, 2026-10-23). In production, keep the shadow runs: they check a new version before it serves anyone.
