Fail-open vs fail-closed

For LLM cost tooling the choice decides what a vendor outage does to your application. Fail-closed enforcement means the vendor's downtime becomes your downtime; fail-open means an outage costs you enforcement, never availability. Neither is universally right — the honest position is to say which one you chose and why.

Capsera chooses fail-open at every layer: if its backend is unreachable, the pre-call budget check is skipped and the call proceeds, and event tracking queues locally rather than blocking. The trade, stated plainly: during such an outage, budgets are not enforced.

By behaviour under fault injection — kill the control plane and observe whether application calls proceed (open) or error (closed), and what happens to the events generated meanwhile.

What is fail-open vs fail-closed?

A fail-open control stops controlling when its infrastructure is unreachable and lets traffic proceed; a fail-closed control blocks traffic instead, trading availability for guarantees. For LLM cost tooling the choice decides what a vendor outage does to your application. Capsera fails open at every layer, so an outage costs you enforcement for that moment, never availability.

• Pre-call enforcement — Pre-call enforcement is a budget check that runs before an LLM request is sent to the provider, so an over-budget call is stopped before it costs anything rather than reported after it has.
• Agentic spend governance — Agentic spend governance is the practice of attributing, budgeting, and enforcing LLM spend at the level of the agents that generate it, rather than at the level of API keys or accounts.
• Runaway agent loop — A runaway agent loop is an agent retrying or re-invoking itself without a terminating condition, generating LLM spend continuously until something outside the loop stops it.