메뉴
Run structured evaluations (evals) for any skill that has an evals/evals.json file. Executes each test case with and without the skill loaded, grades outputs against assertions, and produces grading, benchmark, and report files. Use when asked to eval, test, or validate a skill.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Format Go code and generate GoDoc documentation adhering to Go best practices. Use when prompted "adjust this code", "format code", "sanitize code", "document this code", "add documentation"
Reference knowledge for DDD module composition in this Golang hexagonal architecture project. Provides layer structure, naming conventions, implementation patterns, and design guidelines for domain, application, and infrastructure layers. Use this skill when creating, modifying, analyzing, or reviewing DDD modules, or when any task requires understanding the module architecture, layer boundaries, or component patterns of this codebase.
Run structured evaluations (evals) for any skill that has an evals/evals.json file. Executes each test case twice — once with the skill loaded (with_skill) and once without it (without_skill baseline) — then grades outputs against assertions via a dedicated haiku grader subagent, generates per-eval grading.json, timing.json, benchmark.json, and a feedback.json for human review. Grading and reporting are offloaded to subagents to keep the main context lean. Use when the user says "eval this skill", "run evals for", "test this skill", or provides a skill path and asks to evaluate it. Requires the target skill to have evals/evals.json.
Generate comprehensive unit tests in Go with 100% coverage following specific structural standards. Use when prompted "add unit tests", "add new unit tests", "test this code", "test this", "make unit test", "make test"
Stateless SDD (Spec-Driven Development) pipeline orchestrator. Routes the user through the phases prd -> plan -> implement -> verify -> done, keeping each interactive phase isolated in its own session. Use when: starting a new feature from scratch, checking pipeline status, or advancing to the next phase. Triggers: "/sdd-workflow", "next phase", "advance pipeline", "sdd status".
Create a technical specification plan for a specific feature, from a PRD or a feature description. Use it when the user requests a technical specification plan or needs help creating one. Reads a PRD (from the spec-prd skill or provided by the user), elicits and confirms scope, then produces an approved PLAN.md whose tasks are annotated with dependencies and complexity, following plan-schema.
| name | eval-skill |
| description | Run structured evaluations (evals) for any skill that has an evals/evals.json file. Executes each test case with and without the skill loaded, grades outputs against assertions, and produces grading, benchmark, and report files. Use when asked to eval, test, or validate a skill. |
| execution_profile | {"tier":"speed","strategy":"plan"} |
Load the full skill instructions by reading .agentic/skills/eval-skill/SKILL.md before proceeding with any task.
