latency-budget-architect

What problem does it solve?

This Skill turns an end-to-end latency target into explicit per-hop engineering budgets so teams can own milliseconds instead of guessing with defaults and folklore.

Core Features & Use Cases

• Budget decomposition: Split a user-facing latency promise across edge, gateway, services, databases, queues, and third parties with named ownership.
• Tail-aware planning: Account for fan-out, percentile behavior, serialization overhead, connection setup, queue wait, and retry cost so the arithmetic closes honestly.
• Timeout and retry design: Derive timeouts and retries from the budget rather than from inherited defaults, and add headroom plus violation attribution.
• Use case: A checkout path has a 500ms p95 target and needs budget allocation, timeout derivation, and a rule for new dependencies that want a share of the path budget.

Quick Start

Ask for a per-hop latency budget for your target journey, including headroom, derived timeouts, retry policy, and budget-claim rules for any new dependency.