Restart the service.
Skip the ticket.
Self-service restarts for the services you own, pre-permissioned through Spellbook: you move the moment something needs a kick, and infra keeps exact control over who can touch what.
▸ revivify restart payments-api
The spell, live
Infra writes the policy once. The developer just casts.
- restart:checkout-apiteam:paymentsgranted
- restart:payments-apiteam:paymentsgranted
- restart:search-apiteam:searchgranted
- delete:ledger-db—denied
Pre-authorized once. No ticket, no approval step at cast time — and nothing outside the grant, ever.
What it does
Restart it yourself
Your service is crashlooping. Cast Revivify and it comes back. No ticket, no paging, no waiting.
Pre-permissioned, not pre-approved
Infra grants the scope once; every restart after that is instant, checked against standing policy.
Bounded by the ward
You can only restart what your grant allows: no standing credentials, no reach beyond the policy.
Every cast is on the record
Who restarted what, when, under which scope, written to the audit trail automatically.
How it's warded
Standing policy, not per-restart approval. Least privilege, full audit trail.
For developers
When the service you own falls over, bring it back yourself in seconds. No context switch, no ticket queue.
▸ revivify restart checkout-api ◇ pre-authorized ✓ restart:checkout-api ✔ healthy · back online in 8.4s
For infra
Grant a scope, not a standing key. Developers move inside a policy you set; every cast is logged.
grant team:payments → restart:checkout-api ✓ → restart:payments-api ✓ → delete:ledger-db ✕ denied