What problem does it solve? Coding agents often over-engineer solutions, refactor unrelated code, or declare tasks done without verification. This Skill provides a non-negotiable set of operating principles that govern every coding task, preventing scope creep, premature abstraction, and unverified completions. ## Core Features & Use Cases - Simplicity and Surgical Changes: Enforces YAGNI via a leverage ladder (reuse codebase code, standard library, platform features, existing dependencies before writing new code) and forbids touching code outside the task scope. - Goal-Driven Execution and Definition of Done: Requires verifiable success criteria per task and a strict completion checklist covering format, lint, tests with coverage, judge review, security review, and live browser verification. - Branch and Commit Discipline: Mandates typed branches (feature/, fix/, hotfix/, refactor/, chore/, docs/), conventional commits, and micro-PRs under 12 files and 3000 lines. - Use Case: When starting any non-trivial coding task, the agent restates the goal, lists success criteria, reads target files before editing, and loops through verification until all Definition of Done checks pass. ## Quick Start Ask the agent to implement a small feature and require it to follow the core operating principles including branch discipline and the Definition of Done checklist.
