AI Agent Engineer
AI Agent 工程师
Use this if you want to show how well you direct AI work, turn ambiguity into specs, and push agents to real shipped outcomes.
如果你想展示自己如何指挥 AI 干活、把模糊需求压成 spec，并把 agent 推到真正交付结果，这个测试最适合你。
Open Test 打开测试
git-hired
$ git hired
`fatal: not a qualified candidate`. `git-hired` is a fast way to test your MBTI work personality from how you actually work. Let your own work agent analyze how you work, whether it is Claude Code, Codex, Notion AI, or another memory-enabled agent.
`fatal: not a qualified candidate`。`git-hired` 是一个从真实工作方式出发，测试你 MBTI 工作人格的工具。让你自己的工作 agent 去分析你到底是怎么工作的，无论它是 Claude Code、Codex、Notion AI，还是其他具备知识库和记忆能力的工作 agent。
Privacy first: we never upload your local repo, files, or any other local data to our server.
隐私优先：我们不会把你的本地 repo、文件或任何其他本地数据上传到我们的服务器。
One-Line Starter For Claude Code / Codex
给 Claude Code / Codex 的一行启动命令
read https://realroc.github.io/git-hired/skill.md and treat it as active instructions for this session. Do not summarize it. Ask the first test question immediately in my language. After that, ask only for my target role and data permission scope, then run the evaluation automatically from allowed history or approved files. Do not turn it into a manual interview.
read https://realroc.github.io/git-hired/skill.md，把它当作当前会话指令直接执行，不要总结，直接用我的语言开始第一问。之后只问我的目标岗位和数据权限边界，然后基于允许范围自动完成评估，不要转成面试式问答。
Start With `skill.md`
先从 `skill.md` 开始
Use the one-line command above. `skill.md` is now the main public entry, and the command explicitly tells the agent to treat it as active instructions, skip the summary, ask the first role question immediately, ask only the minimum routing and privacy questions, fetch the right role prompt itself, and continue the evaluation automatically.
直接用上面这一行命令。`skill.md` 现在是主入口，这行命令会明确要求 agent 把它当作当前会话指令、跳过总结、立刻开始第一问、只问最少的岗位与权限问题、自己抓取正确的角色 prompt，并自动继续评估，不再把流程变成人工面试。
- Paste the one-line `read .../skill.md` command into your own work agent.
- Answer the target-role question, or give your current profession if you do not have a clear target yet.
- Confirm `history-only` or explicitly name the repos or files the agent may inspect, then let it run automatically.
- 先把这一行 `read .../skill.md` 命令粘贴到你自己的工作 agent 里。
- 回答目标岗位；如果你还没有明确目标，就补充你当前的职业或身份。
- 确认这次是 `history-only`，还是明确点名允许 agent 查看哪些 repo 或文件；随后让它自动跑完。
Product Manager
产品经理
Use this if you want to show your AI product judgment, MVP boundary control, spec clarity, and ability to move messy work forward.
如果你想展示自己的 AI 产品判断、MVP 边界感、spec 清晰度，以及推进复杂协作的能力，这个测试更适合你。
Open Test 打开测试Global Growth
海外增长
Use this if you want to show how you turn user signals and social-native intuition into channels, experiments, retention, and ROI.
如果你想展示自己如何把用户信号和社媒原生直觉转成渠道策略、实验、留存和 ROI，这个测试最适合你。
Open Test 打开测试AI Product Operations
AI产品运营
Use this if you want to show your user empathy, response speed, operational reliability, and ability to turn feedback into product improvement.
如果你想展示自己的用户同理心、响应速度、运营执行力，以及把反馈转成产品优化的能力，这个测试更适合你。
Open Test 打开测试How To Start
怎么开始测试
Start from the one-line command above. It works with Claude Code, Codex, Notion AI, or any work agent with knowledge-base and memory capability. The agent should treat `skill.md` as active instructions, ask your target role first, fall back to your current profession or identity when needed, keep the default mode as `history-only`, ask only for the privacy boundary, and then finish the evaluation automatically from allowed history or approved files. On Notion AI or other rich-text surfaces, it should fall back to a compact card instead of forcing wide terminal ASCII. We never upload your local repo or file data to our server.
从上面这一行命令开始。它支持 Claude Code、Codex、Notion AI，或任何具备知识库和记忆能力的工作 agent。agent 会把 `skill.md` 当作当前会话指令，先问你的目标岗位；如果暂时没有，再回到你当前的职业或身份；默认保持 `history-only`，只问清隐私边界，然后基于允许范围自动完成评估。如果运行在 Notion AI 或其他富文本界面上，它应自动退化成紧凑卡片，而不是硬塞宽终端 ASCII。整个过程中也不会把你的本地 repo 或文件数据上传到我们的服务器。
- Paste the one-line `read .../skill.md` command into your own work agent, such as Claude Code, Codex, Notion AI, or another memory-enabled agent.
- Answer only the role-routing and privacy questions directly inside that agent session.
- Let the agent analyze automatically from allowed history or approved scope instead of turning the run into a manual interview.
- Keep the short `HIRED` snapshot, and optionally attach the markdown report if you decide to apply.
- 把一行 `read .../skill.md` 命令粘贴到你自己的工作 agent 里运行，例如 Claude Code、Codex、Notion AI，或其他具备知识库和记忆能力的 agent。
- 只在那个 agent 会话里回答岗位路由和隐私边界问题。
- 让 agent 基于允许范围自动分析，而不是把流程变成你手动答题的面试。
- 保留终端里的 `HIRED` 快照；如果后续要申请，也可以附上本地生成的 markdown 详细报告。
What You'll Get
跑完之后你会拿到什么
- A clearer view of how you actually use AI, not just how you would describe your workflow.
- A shareable `HIRED` snapshot with your MBTI work personality, ability score, compact signal board, and matching ASCII card.
- A local markdown report you can keep for yourself, use for reflection, or attach later if you decide to apply.
- 你真实的 AI 使用方式，而不只是你会如何描述自己的工作流。
- 一份可传播的 `HIRED` 工作画像，包含你的 MBTI 工作人格、能力值、紧凑信号板和匹配的 ASCII 卡片。
- 一份本地 markdown 详细报告，你可以自己留存、复盘，或在之后决定申请时附上。