agentic-engineering

What problem does it solve? Engineering teams delegating implementation work to AI agents often lack a disciplined process for defining completion criteria, verifying output quality, and controlling model costs, leading to unreliable results and wasted tokens. ## Core Features & Use Cases - Eval-First Execution: Define capability and regression evals before implementation, capture baseline failure signatures, and compare deltas after changes. - Task Decomposition: Break work into 15-minute, independently verifiable units, each with a single dominant risk and a clear done condition. - Cost-Aware Model Routing: Assign Haiku to narrow edits, Sonnet to implementation, and Opus to architecture and root-cause analysis, escalating tiers only on clear reasoning gaps. - Use Case: A team lead asks an agent to refactor a multi-module service; the skill structures the work into verifiable units, routes each unit to the right model tier, and validates results with regression evals. ## Quick Start Use the agentic-engineering skill to plan and execute my next refactoring task with eval-first verification and tiered model routing.