availability

A classic warm-up that rewards precision: what state the server keeps, why spoofed SYNs cost the attacker nothing, and how SYN cookies move that state into the sequence number.

Pinning trades availability for a narrower trust set. The answer that scores compares each pin target by what it survives, and uses HPKP's retirement as evidence of what goes wrong.

A 2 Tbps flood and one expensive API call per second are both denial of service. The answer that scores matches each attack to the place where it can be stopped, and knows which ones only a provider can absorb.

The internet's routing system still largely believes what networks announce. The strong answer explains why a more specific prefix wins, what RPKI signs, and why path manipulation survives it.

Small requests that trigger large work are the cheapest denial of service. The scored answer finds the multipliers, bounds every one, and adds the AI-era ones.

An AI request can cost a thousand times more than a web request, so volumetric thinking alone fails. The scored answer layers network, request and cost controls, and protects the GPU tier from ever seeing junk.

A fallback changes the real request path. Review recipient eligibility, state compatibility and uncertain actions before trying another endpoint.

Review caller-controlled options by their actual information, resource and capability effects. Build a bounded product schema instead of forwarding arbitrary configuration.

An easy question with a revealing follow-up. The answer that scores limits cost at every layer, including the agent loop, not just requests per minute.

There is no single answer, and saying so is the start. The scored answer decides per check from what its failure would let through, and designs the latency budget so the choice is rarely forced.

Bound work where it is dispatched, account for concurrent branches and retries, and keep uncertain operations reserved. Choose limits from workload evidence and operational capacity, then preserve useful progress when a bound is reached.

Separate false-information spread, borrowed authority and runaway work. Put shared limits and stop checks at the propagation paths, then recover without repeating completed effects.