primer/agent-eval

최근 기록된 소스 활동: 2026년 10월 24일 16:16

SkillsMP 카탈로그 업데이트: 2026년 10월 29일 18:30

수집된 skills: 7

GitHub 스타: 2
GitHub 포크: 0

AI 어시스턴트로 설치

이 프롬프트를 복사해 사용 중인 AI 어시스턴트에 입력하세요.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/primer/agent-eval.

저장소 README

README 원본 파일: README.md

README 원본 보기

README 전체 펼치기

# agent-eval > An open source library for evaluating model performance across different tasks. This project contains the [`@primer/agent-eval`](./packages/agent-eval) library and CLI for evaluating agent performance across different tasks. This is useful for establishing [benchmarks](./docs/benchmarks.md) for the capabilities of your project (like a design system) or for running [experiments](./docs/experiments.md) to understand the best way to improve model outcomes for the task at hand. To do so, we create a variety of [scenarios](./docs/scenarios.md) that describe the task, the starting workspace, and the checks or judges that evaluate the result. These scenarios are then used in benchmarks and experiments to evaluate model performance. To learn more about how to use this library, visit [`@primer/agent-eval`](./packages/agent-eval) or install the [agent-eval skill](./skills/agent-eval/SKILL.md) to get started. ## Agent skills Install the agent-eval skill to help your agent set up and run evaluations: ```sh npx skills add primer/agent-eval --skill agent-eval ``` The [skill](./skills/agent-eval/SKILL.md) points to the docs included in your installed `@primer/agent-eval` package, so guidance stays aligned with that version. The skill and runtime package are installed separately. You can also skip installing the skill and run `npx agent-eval agent setup` from your project's root to add this guidance automatically. Or, after installing the package, add this one line to your project's `AGENTS.md`: > When working with agent-eval, first read `node_modules/@primer/agent-eval/docs/AGENTS.md`. The [package docs entrypoint](./packages/agent-eval/docs/AGENTS.md) links to getting-started, evaluation methodology, domain model, and CLI guides. Install the [agent-eval-investigate skill](./skills/agent-eval-investigate/SKILL.md) to investigate experiment or benchmark results, explain observed model behavior (such as why a tool was or was not called), and suggest setup improvements: ```sh npx skills add primer/agent-eval --skill agent-eval-investigate ``` Use `/agent-eval-investigate` with a result bundle, the relevant scenario or trial, and the behavior you want to understand. The skill can also offer to apply a proposed change and rerun a focused comparison with your approval. ## Contributing We love collaborating with folks inside and outside of GitHub and welcome contributions! If you're interested, check out our [contributing docs](.github/CONTRIBUTING.md) for more info on how to get started. ## License Licensed under the [MIT License](./LICENSE).

저장소 README를 불러오는 중…

이 저장소의 skills

분류 대기 중

skill

직업 분류

설명

업데이트

architecture

미분류

Use when changing orchestration, lifecycle, retry or cleanup behavior, resource ownership, or dependency boundaries in agent-eval. Do not use for isolated parsing, formatting, documentation, or UI changes that leave these boundaries unchanged.

원문 언어: 영어

2026-10-24

testing

미분류

Use when adding, updating, or reviewing tests in agent-eval, choosing coverage, testing lifecycle or concurrency behavior, or isolating external dependencies. Do not use for scenario grading or changes that do not involve repository tests.

원문 언어: 영어

2026-10-24

changesets

미분류

Use when changing public APIs or behavior, deprecating usage, recommending replacements, or preparing releases; maintain changesets and versioned migration guides.

원문 언어: 영어

2026-10-21

agent-eval

미분류

Use when installing or using @primer/agent-eval, designing agent evaluations, creating benchmarks, experiments, scenarios, checks, judges, or treatments, and running, planning, sharding, or interpreting evaluations with the agent-eval CLI.

원문 언어: 영어

2026-10-20

ui

미분류

Use to understand how to build and verify UI-related work

원문 언어: 영어

2026-10-19

agent-eval-investigate

미분류

Use when investigating model behavior in agent-eval experiment or benchmark results, explaining why an agent did or did not call a tool, follow an instruction, or use a skill, and proposing setup improvements with optional iterative reruns.

원문 언어: 영어

2026-10-09

turborepo

미분류

Turborepo monorepo build system guidance. Triggers on: turbo.json, task pipelines, dependsOn, caching, remote cache, the "turbo" CLI, --filter, --affected, CI optimization, environment variables, internal packages, monorepo structure/best practices, and…

원문 언어: 영어

2026-09-16

수집된 skill 7개 중 7개를 표시합니다.