菜单
Prepare for job interviews with tailored questions, STAR answers, and company research. Builds interactive web apps for practice.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Design static ad creatives for social media and display advertising campaigns.
Source and evaluate candidates with job analysis, search strategies, specific candidate profiles, and outreach templates.
Draft emails, manage calendar scheduling, prepare meeting agendas, and organize productivity
Create brand identity kits with color palettes, typography, logo concepts, and brand guidelines.
Perform competitive market analysis with feature comparisons, positioning, and strategic recommendations.
Create social media posts, newsletters, and marketing content calibrated to your voice and platform.
| name | interview-prep |
| description | Prepare for job interviews with tailored questions, STAR answers, and company research. Builds interactive web apps for practice. |
Prepare for interviews with company-specific research, behavioral story banks, technical frameworks, and salary negotiation scripts. Always delivers as an interactive web app.
Before doing anything else, ask the user:
All three modes are built as web apps deployed as artifacts. Do not output raw markdown — always build an interactive app.
Before building anything, use webSearch to gather real intel. This data populates whichever mode the user chose.
| Source | Query | What you get |
|---|---|---|
| Glassdoor | site:glassdoor.com [Company] interview questions [role] | Actual questions asked, by round. Filter to last 6 months. |
| Blind | site:teamblind.com [Company] interview OR [Company] onsite | Unfiltered loop structure, which rounds matter, bar-raiser tells, comp bands |
| levels.fyi | site:levels.fyi [Company] | Real comp by level + location. Critical for negotiation anchoring. |
| LeetCode Discuss | site:leetcode.com/discuss [Company] [role] | Tagged coding problems actually asked |
[Company] [team name] → recent posts from hiring manager | What they're shipping, what they celebrate, vocabulary they use | |
| Eng blog / Newsroom | [Company] engineering blog | System design context. Mentioning their published architecture in an interview is a strong signal. |
For Amazon specifically: each interviewer is assigned 1-3 Leadership Principles to probe. Map stories to LPs before the loop.
Build a React web app that uses the OpenAI Realtime API for voice-to-voice interview simulation. The app should:
// OpenAI Realtime API connection pattern
const pc = new RTCPeerConnection();
const audioEl = document.createElement("audio");
audioEl.autoplay = true;
// Add local audio track for user's microphone
const ms = await navigator.mediaDevices.getUserMedia({ audio: true });
pc.addTrack(ms.getTracks()[0]);
pc.ontrack = (e) => { audioEl.srcObject = e.streams[0]; };
// Connect to OpenAI Realtime
const tokenRes = await fetch("/api/realtime-session", { method: "POST" });
const { client_secret } = await tokenRes.json();
const offer = await pc.createOffer();
await pc.setLocalDescription(offer);
const sdpRes = await fetch("https://api.openai.com/v1/realtime?model=gpt-4o-realtime-preview", {
method: "POST",
headers: {
Authorization: `Bearer ${client_secret.value}`,
"Content-Type": "application/sdp",
},
body: offer.sdp,
});
await pc.setRemoteDescription({ type: "answer", sdp: await sdpRes.text() });
The backend route (/api/realtime-session) creates an ephemeral token via:
const r = await fetch("https://api.openai.com/v1/realtime/sessions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.OPENAI_API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
model: "gpt-4o-realtime-preview",
voice: "verse",
instructions: `You are an interviewer for [Role] at [Company]. Ask behavioral and technical questions one at a time. After the candidate answers, give brief feedback, then move to the next question. Questions: [insert researched questions from Step 1]`,
}),
});
const data = await r.json();
// Return data.client_secret to the frontend
Build a React chat app that simulates a text-based interview. The app should:
System prompt for the interviewer AI:
You are interviewing a candidate for [Role] at [Company]. Ask questions one at a time from this list: [researched questions]. After each answer, give brief, constructive feedback focusing on: specificity, use of "I" vs "we", quantified results, and STAR structure. Then ask the next question. Be professional but conversational.
Build a polished single-page React app with these sections:
STAR is the baseline. Know the variants:
Build a 6-story matrix, not a script per question. One well-told story covers 3-4 question variants.
| Category | Prepare 1 story each | Maps to questions like |
|---|---|---|
| Leadership | Stepped up without authority | "Influenced without authority," "led a project" |
| Conflict | Disagreed with manager/peer, resolved | "Difficult coworker," "disagree and commit" |
| Failure | Owned a mistake, quantify damage + fix | "Project that failed," "missed a deadline" |
| Ambiguity | Decided with incomplete info | "Moved fast," "prioritized under uncertainty" |
| Impact | Your single biggest measurable win | "Most proud of," "biggest accomplishment" |
| Scope creep | Did more than asked | "Exceeded expectations," "ownership" |
"Tell me about yourself": Present (current role + one win) → Past (how you got here, 1 pivot) → Future (why this role is the logical next step). 90 seconds.
System design — RADIO:
Coding pattern recognition: ~75% of LeetCode-style questions reduce to: sliding window, two pointers, BFS/DFS, binary search on answer, heap for top-K, DP (1D/2D), union-find, monotonic stack.