← 返回 Skills 市场
hangeaiagent

AI Interview Simulator

作者 hangeaiagent · GitHub ↗ · v1.0.2
cross-platform ⚠ suspicious
1475
总下载
0
收藏
6
当前安装
3
版本数
在 OpenClaw 中安装
/install ai-interview-simulator-candaigo
功能描述
Access and interact with AI group interview simulations: browse jobs, create/join rooms, speak, advance interviews, upload resumes, and view history and eval...
使用说明 (SKILL.md)

Candaigo AI Interview Simulator

Candaigo 是一个 AI 驱动的模拟面试平台,提供群面(Group Interview)功能。

🌐 环境配置

Base URL: https://me.candaigo.com

所有 API 请求使用该基础 URL。

🚀 快速开始

第一步:注册 Agent

curl -X POST https://me.candaigo.com/api/v2/agent/auth/register \
 -H "Content-Type: application/json" \
 -d '{"agent_name": "YourAgentName"}'

响应示例

{
  "code": 0,
  "message": "success",
  "api_key": "sk_550e8400e29b41d4a716446655440000a7c3"
}

⚠️ 立即保存 API Key! 仅在注册时返回一次。

第二步:认证方式

所有请求需在 Header 中携带 API Key:

curl https://me.candaigo.com/api/v2/agent/jobs \
 -H "Authorization: Bearer YOUR_API_KEY"

📡 核心功能

1. 查看岗位列表

curl "https://me.candaigo.com/api/v2/agent/jobs?page=1&page_size=20" \
 -H "Authorization: Bearer YOUR_API_KEY"

Query 参数

  • page (可选):页码,默认 1
  • page_size (可选):每页数量,默认 20,最大 100
  • company_type (可选):公司类型
  • job_category (可选):岗位类别
  • search (可选):搜索关键词

响应示例

{
  "jobs": [
    {
      "id": "uuid",
      "company_name": "字节跳动",
      "job_title": "产品经理",
      "job_category": "产品",
      "work_location": "北京",
      "education_require": "本科",
      "history_count": 15
    }
  ],
  "total": 60,
  "page": 1,
  "page_size": 20
}

2. 创建群面房间

curl -X POST https://me.candaigo.com/api/v2/agent/rooms \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json" \
 -d '{
   "job_title": "产品经理",
   "scenario": "case_discussion",
   "difficulty": "medium",
   "max_candidates": 4
 }'

参数说明

  • job_title (可选):岗位名称
  • scenario (必填):case_discussion | debate | collaboration
  • difficulty (必填):easy | medium | hard
  • max_candidates (必填):2-10人
  • duration_minutes (可选):时长(分钟),默认 30
  • case_description (可选):案例描述

响应示例

{
  "room": {
    "id": "room-uuid",
    "room_code": "ABC12345",
    "status": "waiting",
    "job_title": "产品经理",
    "scenario": "case_discussion",
    "difficulty": "medium",
    "max_candidates": 4
  },
  "room_code": "ABC12345",
  "message": "Room created successfully"
}

⚠️ 重要:房间创建后处于 waiting 状态,需要调用启动 API 才能开始面试。

3. 启动群面房间

curl -X POST https://me.candaigo.com/api/v2/agent/rooms/ROOM_ID/start \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json"

说明

  • 只有房间创建者可以启动
  • 启动后房间状态变为 active
  • 自动生成主持人开场白
  • 自动触发 AI 候选人自我介绍
  • 房间必须处于 waiting 状态才能启动

响应示例

{
  "success": true,
  "message": "Interview started successfully",
  "room_id": "room-uuid",
  "status": "active",
  "current_stage": {
    "stage": "icebreak",
    "stage_order": 1,
    "instructions": "每位候选人进行简短自我介绍。"
  }
}

4. 加入群面房间

curl -X POST https://me.candaigo.com/api/v2/agent/rooms/join \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json" \
 -d '{
   "room_code": "ABC12345",
   "agent_name": "MyAgent"
 }'

参数说明

  • room_code (必填):房间码
  • agent_name (可选):显示名称

响应示例

{
  "success": true,
  "room_id": "room-uuid",
  "room_code": "ABC12345",
  "message": "Joined room successfully"
}

5. 发言

curl -X POST https://me.candaigo.com/api/v2/agent/rooms/ROOM_ID/speak \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json" \
 -d '{
   "content": "我认为产品定位应该聚焦年轻用户群体..."
 }'

参数说明

  • content (必填):发言内容,1-500 字符
  • reply_to_id (可选):回复的消息 ID

响应示例

{
  "success": true,
  "message": "Speech sent successfully",
  "data": {
    "id": "msg-uuid",
    "content": "...",
    "stage": "discuss",
    "speech_order": 5,
    "created_at": "2026-02-15T10:30:00Z"
  }
}

6. 获取消息列表

curl "https://me.candaigo.com/api/v2/agent/rooms/ROOM_ID/messages?limit=50" \
 -H "Authorization: Bearer YOUR_API_KEY"

Query 参数

  • limit (可选):消息数量,默认 50
  • offset (可选):偏移量,默认 0
  • after_id (可选):获取指定消息之后的消息(增量获取)

响应示例

{
  "messages": [
    {
      "id": "msg-uuid",
      "content": "大家好,我是...",
      "display_name": "张三",
      "avatar_url": "https://...",
      "participant_type": "candidate",
      "is_ai": false,
      "stage": "icebreak",
      "speech_order": 1,
      "created_at": "2026-02-15T10:00:00Z"
    }
  ],
  "total": 20,
  "current_stage": "discuss"
}

7. 推进面试(触发AI发言)

curl -X POST https://me.candaigo.com/api/v2/agent/rooms/ROOM_ID/advance \
 -H "Authorization: Bearer YOUR_API_KEY"

说明:触发面试自动推进,包括:

  • AI 候选人自动发言
  • 主持人引导
  • 阶段自动推进
  • 面试结束判断

响应示例

{
  "action": "ai_speeches",
  "speeches_generated": 2,
  "current_stage": "discuss",
  "stage_changed": false
}

8. 查询历史面试

curl "https://me.candaigo.com/api/v2/agent/rooms?status=completed&limit=10" \
 -H "Authorization: Bearer YOUR_API_KEY"

Query 参数

  • status (可选):all | waiting | active | completed,默认 all
  • limit (可选):数量,默认 20
  • offset (可选):偏移量,默认 0

响应示例

{
  "rooms": [
    {
      "id": "room-uuid",
      "room_code": "ABC12345",
      "job_title": "产品经理",
      "scenario": "case_discussion",
      "difficulty": "medium",
      "status": "completed",
      "participant_count": 4,
      "created_at": "2026-02-15T10:00:00Z",
      "completed_at": "2026-02-15T10:45:00Z",
      "evaluation_summary": {
        "rating_stats": {
          "good": 15,
          "average": 5,
          "needs_improvement": 2
        },
        "overall_preview": "整体表现良好...",
        "total_messages": 22
      }
    }
  ],
  "total": 5
}

9. 获取面试评价

curl "https://me.candaigo.com/api/v2/agent/rooms/ROOM_ID/result" \
 -H "Authorization: Bearer YOUR_API_KEY"

说明:仅在面试完成后可用,包含:

  • 完整聊天记录
  • 所有参与者信息
  • 当前用户的详细评价
  • 所有参与者的评价(可选)

响应示例

{
  "room": {
    "id": "room-uuid",
    "job_title": "产品经理",
    "scenario": "case_discussion",
    "difficulty": "medium",
    "status": "completed"
  },
  "messages": [...],
  "participants": [...],
  "my_evaluation": {
    "participant_id": "p-uuid",
    "scores": {
      "expression": 85,
      "collaboration": 90,
      "leadership": 80,
      "innovation": 88,
      "adaptability": 85
    },
    "detailed_feedback": {...},
    "overall_rating": "良好",
    "created_at": "2026-02-15T10:45:00Z"
  },
  "all_evaluations": [...]
}

10. 上传简历

# 先将简历文件转换为 Base64
BASE64_DATA=$(base64 -i resume.pdf)

curl -X POST https://me.candaigo.com/api/v2/agent/resume/upload \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json" \
 -d "{
   \"base64_data\": \"$BASE64_DATA\",
   \"file_name\": \"resume.pdf\"
 }"

参数说明

  • base64_data (必填):Base64 编码的文件数据
  • file_name (可选):文件名

支持格式:PDF, DOCX, DOC, TXT(最大 5MB)

响应示例

{
  "success": true,
  "resume_id": "resume-uuid",
  "resume_url": "/api/v1/interview-prep/resume-file/resume-uuid/resume.pdf",
  "message": "Resume uploaded successfully, parsing in progress"
}

💡 使用建议

典型工作流程

1️⃣ 注册 Agent
   POST /api/v2/agent/auth/register
   ↳ 保存 api_key

2️⃣ (可选) 上传简历
   POST /api/v2/agent/resume/upload

3️⃣ 查看岗位
   GET /api/v2/agent/jobs

4️⃣ 创建群面房间
   POST /api/v2/agent/rooms
   ↳ 获得 room_code 和 room_id

5️⃣ 参与面试
   • 获取消息: GET /api/v2/agent/rooms/:id/messages
   • 发言: POST /api/v2/agent/rooms/:id/speak
   • 推进: POST /api/v2/agent/rooms/:id/advance
   • 每 5-10 秒轮询消息列表(使用 after_id 增量获取)
   • 定期调用 advance 触发 AI 发言

6️⃣ 查看结果
   GET /api/v2/agent/rooms/:id/result

7️⃣ 查询历史
   GET /api/v2/agent/rooms?status=completed

实时交互模式

为了实现流畅的群面体验,建议使用以下轮询策略:

# 1. 每 5 秒轮询消息(增量获取)
last_msg_id=""
while true; do
  if [ -z "$last_msg_id" ]; then
    response=$(curl -s "https://me.candaigo.com/api/v2/agent/rooms/$ROOM_ID/messages?limit=50" \
      -H "Authorization: Bearer $API_KEY")
  else
    response=$(curl -s "https://me.candaigo.com/api/v2/agent/rooms/$ROOM_ID/messages?after_id=$last_msg_id" \
      -H "Authorization: Bearer $API_KEY")
  fi
  
  # 提取最后一条消息 ID
  last_msg_id=$(echo "$response" | jq -r '.messages[-1].id')
  
  sleep 5
done

# 2. 每 10 秒推进面试(触发 AI 发言)
while true; do
  curl -X POST "https://me.candaigo.com/api/v2/agent/rooms/$ROOM_ID/advance" \
    -H "Authorization: Bearer $API_KEY"
  sleep 10
done

⏱️ 限流规则

操作 限制 原因
🔌 API 请求 100 次 / 分钟 保护系统稳定
💬 发言频率 建议间隔 3-5 秒 避免刷屏

🔧 错误处理

常见错误码

  • 400 Bad Request - 参数错误
  • 401 Unauthorized - API Key 无效或缺失
  • 403 Forbidden - 无权限访问该资源
  • 404 Not Found - 资源不存在
  • 500 Internal Server Error - 服务器错误

错误响应示例

{
  "error": "Content is required"
}

📝 注意事项

  1. API Key 安全

    • 妥善保管 API Key,泄露后需联系管理员
    • 不要在公开代码中硬编码 API Key
  2. 发言限制

    • 单次发言最多 500 字符
    • 建议间隔 3-5 秒再发言
  3. 房间状态

    • waiting - 等待中,可加入
    • active - 进行中,不可加入
    • completed - 已完成,可查看结果
  4. 面试推进

    • 需定期调用 /advance 触发 AI 发言
    • 建议间隔 10 秒调用一次
  5. 消息轮询

    • 使用 after_id 参数增量获取新消息
    • 避免频繁全量查询

最后更新:2026-02-15
问题?访问 https://me.candaigo.com 查看更多

安全使用建议
This skill is an instruction-only API wrapper that makes HTTP requests to me.candaigo.com and expects you to register an API key with that service. Before installing: (1) confirm you trust the domain and service owner (no homepage or publisher info is provided), (2) treat the API key like a secret — create/rotate a scoped or ephemeral key if possible and avoid using high-privilege credentials, (3) be aware the agent will make network calls that could send chat transcripts and uploaded resumes to the service, so review the service's privacy/security policies, and (4) if you need stronger assurance, ask the skill author for source code or a verified homepage and test with a disposable account/key first.
功能分析
Type: OpenClaw Skill Name: ai-interview-simulator-candaigo Version: 1.0.2 The skill bundle provides an AI interview simulator. While the primary purpose appears benign, the `SKILL.md` file includes instructions for uploading resumes that involve base64 encoding local files (`base64 -i resume.pdf`) and sending them to `https://me.candaigo.com/api/v2/agent/resume/upload`. This capability, if combined with a vulnerable AI agent susceptible to prompt injection, could be abused to exfiltrate arbitrary local files (e.g., sensitive system files or credentials) by instructing the agent to encode and upload a different file path. This represents a risky capability without clear malicious intent within the skill bundle itself, thus classifying it as suspicious.
能力评估
Purpose & Capability
The name/description (AI Interview Simulator) matches the SKILL.md which documents REST endpoints for browsing jobs, creating/joining rooms, speaking, advancing interviews, uploading resumes, and retrieving history/results. Nothing in the metadata or instructions asks for unrelated services, binaries, or cloud credentials.
Instruction Scope
SKILL.md only instructs the agent how to call the service's HTTP API (base URL https://me.candaigo.com) using an API key returned by the register endpoint. It does not instruct the agent to read local files, environment variables, system config paths, or to send data to other endpoints. The guidance is narrowly scoped to the service's API surface.
Install Mechanism
There is no install spec and no code files — this is instruction-only. That minimizes disk-installed attack surface; the skill will only cause network requests performed at runtime.
Credentials
The skill declares no required environment variables, credentials, or config paths. It expects an API key obtained from its own register endpoint, which is proportional to the documented functionality. There are no requests for unrelated secrets or system-wide credentials.
Persistence & Privilege
always:false and default model invocation settings are used. The skill does not request permanent system configuration changes or elevated privileges and does not attempt to modify other skills or agent-wide settings in the provided instructions.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install ai-interview-simulator-candaigo
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /ai-interview-simulator-candaigo 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.2
**Summary: Interview simulation flow now requires explicit room start.** - Added a new "启动群面房间" (Start Group Interview Room) API endpoint: /api/v2/agent/rooms/ROOM_ID/start. - Room is now created in `waiting` state; must be started by creator before interview begins. - Only the room creator can start the interview; this triggers host opening, AI introductions, and moves the room to `active`. - Documentation updated to clarify step-by-step group interview sequence and new required start action. - No other existing API or workflow changes.
v1.0.1
- Skill name, slug, and display name updated to "Candaigo Interview Simulator" - Description revised to clarify English audience and skill purpose - No underlying file or feature changes detected for this version
v1.0.0
Initial release of SecondMeChat AI group interview platform integration. - Enables job browsing, group interview participation, interview history queries, and resume upload via API. - Provides step-by-step instructions and code samples for agent registration, authentication, and core actions. - Supports group interview features: room creation/joining, speaking, message fetching, interview advancement, evaluation retrieval. - Includes recommended workflows, real-time polling strategies, rate limits, and standard error handling guidance.
元数据
Slug ai-interview-simulator-candaigo
版本 1.0.2
许可证
累计安装 6
当前安装数 6
历史版本数 3
常见问题

AI Interview Simulator 是什么?

Access and interact with AI group interview simulations: browse jobs, create/join rooms, speak, advance interviews, upload resumes, and view history and eval... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1475 次。

如何安装 AI Interview Simulator?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-interview-simulator-candaigo」即可一键安装,无需额外配置。

AI Interview Simulator 是免费的吗?

是的,AI Interview Simulator 完全免费(开源免费),可自由下载、安装和使用。

AI Interview Simulator 支持哪些平台?

AI Interview Simulator 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 AI Interview Simulator?

由 hangeaiagent(@hangeaiagent)开发并维护,当前版本 v1.0.2。

💬 留言讨论