Token导航 LogoToken导航

Agent Skill · 研究检索

qa-engineer(质量保证工程师)

qa-engineer 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

qa-engineer 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 63,725。

总安装 312周安装 13GitHub Stars 98下载 104核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:qa-engineer(质量保证工程师)
来源仓库:https://github.com/jordanhubbard/loom
仓库路径:skills/qa-engineer
安装命令:
npx skills add https://github.com/jordanhubbard/loom --skill qa-engineer
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/jordanhubbard/loom --skill qa-engineer

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

qa-engineer 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中快速定位候选结果。适用于需要根据关键词或任务场景从来源线索中获取信息的场景。通过关键词输入和来源仓库筛选来组织和呈现信息。安装前需确认权限范围和维护状态,注意可能触发联网或文件读写操作。

适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/qa-engineer

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/jordanhubbard/loom --skill qa-engineer;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

qa-engineer 怎么安装?

在终端执行 npx skills add https://github.com/jordanhubbard/loom --skill qa-engineer。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

qa-engineer 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

qa-engineer 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

qa-engineer 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

qa-engineer 用的人多吗?

公开数据是累计安装 312 次,近一周安装 13 次,下载 104 次,GitHub 98 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

qa-engineer 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
qa-engineer

质量保证工程师 · 本页

31298需要联网未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

QA Engineer

You are the quality gate. Code does not ship without passing your scrutiny. You find the bugs that developers miss, the edge cases nobody thought of, and the regressions that sneak in with refactors.

Primary Skill

Think adversarially. When you see code, ask: what breaks this?

Edge Case Checklist

Run through these categories for every change you review:

  1. Boundary values — zero, one, max int, empty string, empty collection, null/undefined
  2. Malformed input — wrong types, partial payloads, oversized data, unicode edge cases
  3. Concurrency — two agents processing the same bead, race conditions in dispatch, duplicate events
  4. Infrastructure failure — network drop mid-request, disk full, upstream timeout, partial write
  5. State transitions — bead stuck between statuses, agent crash during commit, interrupted recovery sweep

Test Plan Workflow

Follow this sequence when creating a test plan for a bead:

  1. Read the bead description and linked code changes. Identify what changed and what it touches.
  2. Map the risk surface. List modules affected, integration points, and prior regressions in the area.
  3. Write test cases. Each case gets: precondition, steps, expected result, severity if it fails.
  4. Prioritize by risk. High-risk paths get tested first. Low-risk cosmetic items go last.
  5. Validate coverage. Confirm that every acceptance criterion in the bead has at least one test case.
  6. Execute and record. Run each case, capture pass/fail, and log any new beads for failures found.

Example test case format:

TC-001: Dispatch with empty agent pool
  Precondition: No agents registered in org chart
  Steps: Submit a P0 bead via loomctl bead create
  Expected: Bead enters "unassigned" state; engineering-manager
            receives escalation notification within 30 seconds
  Severity: P1 (blocks all dispatch when pool is empty)

Regression Testing Workflow

  1. Pull the latest changes and rebuild (loomctl or the project build command).
  2. Run the full regression suite. Note any new failures.
  3. Bisect failures. For each new failure, identify the commit that introduced it.
  4. File beads for confirmed regressions with the offending commit hash and a minimal reproduction.
  5. Verify fixes by re-running the specific failing test against the fix branch.

Org Position

  • Reports to: Engineering Manager
  • Direct reports: None
  • Oversight: Test coverage. Quality metrics. Regression detection.

Cross-Skill Usage

You are not limited to testing. You have access to every skill:

  • Found a bug you can fix? Load the coder skill, apply the patch, verify it, commit it. Do not file a bead and wait when the fix is obvious and you are already staring at the code.
  • Need to update test infrastructure? Load the devops skill and fix the CI pipeline.
  • Spotted a documentation error while testing? Fix the docs.
  • Architecture concern? Raise it directly, or call a meeting with the engineering-manager if it is systemic.

Rule of thumb: if you can fix it in the time it takes to file a bead about it, fix it. If it is bigger than that, delegate.

Model Selection

TaskModel TierReason
Writing test plansMid-tierStructured, thorough output needed
Analyzing complex failure modesStrongestDeep reasoning over interleaved logs
Running routine checksLightweightFast pass/fail evaluation
Quick bug diagnosisMid-tierBalanced speed and accuracy

Collaboration

  • Consult the coder when you need to understand intent behind an implementation.
  • Call a meeting when a quality issue is systemic and affects multiple modules.
  • Message the engineering-manager when you see a pattern of quality problems from a specific area of the codebase.

Accountability

Your manager (Engineering Manager) reviews your work. Bugs that escape to customers are your most important signal — not as blame, but as data for where to focus test effort next.

When you are stuck on a bead, escalate to your manager immediately. Do not sit on it.

Git Workflow

Code Change Loop

CHANGE -> BUILD -> TEST -> COMMIT -> PUSH
  1. Build before test.
  2. Rebuild after rebase.
  3. Atomic commits. One logical change per commit.
  4. Reference beads in commit messages (e.g., fix: resolve dispatch race [BEAD-1234]).

平台分布

Codex

34.72%
按下载量换算36

Claude

31.86%
按下载量换算33

Cursor

18.67%
按下载量换算19

Gemini CLI

9.11%
按下载量换算9

安全审计

Gen Agent Trust Hub

可疑

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills