Career Manager — Job Pipeline
Automates job search: finds roles, tracks applications, tailors resumes, preps for interviews, and manages follow-ups.
Data File: data/applications.json
CODEBLOCK0
Resume Tailoring
When user shares a job description:
- 1. Parse key requirements (must-have vs nice-to-have)
- Map each requirement to user's experience (read
profile/career.md) - Suggest bullet point rewrites emphasizing relevant experience
- Flag gaps and suggest how to address in cover letter
- Rate overall match: "You match X/Y requirements strongly, Z partially, N gaps"
Interview Prep
When interview is scheduled:
- 1. Web search: recent company news, product launches, tech blog
- Research interviewer if name provided
- Generate likely questions (technical, behavioral STAR format, system design)
- Prepare talking points per project
- Suggest questions user should ask
- Send prep package 24h before
Follow-Up Management
- - 5 business days after apply, no response → draft follow-up email
- After phone screen → draft thank-you within 24h
- After technical → detailed thank-you referencing discussion
- After onsite → personalized thank-you per interviewer
- Track ghosting patterns
Application Updates via Natural Language
- - "heard back from [company]" → prompt for details, update status
- "got rejected from [company]" → update to rejected, log reason
- "have a phone screen with [company] next Tuesday" → update status, schedule prep
- "got an offer!" → celebrate, then help evaluate
Instructions
- 1. Always check
data/applications.json before suggesting roles (avoid duplicates) - Update JSON immediately after any career conversation
- Be strategic — quality > quantity
- Help spot patterns: what types of roles respond? What keywords work?
- If <10% response rate after 20 apps, reassess approach
- For interviews, always research first — never send generic prep
职业经理 — 求职流程
自动化求职流程：寻找职位、追踪申请、定制简历、准备面试、管理跟进。
数据文件：data/applications.json
json
{
applications: [
{
id: app_001,
company: 示例公司,
role: 软件工程师,
url: ,
status: 已申请,
applied_date: 2026-02-01,
source: 领英,
contact: null,
notes: ,
followupdate: 2026-02-08,
interviews: [],
outcome: null
}
],
stats: { totalapplied: 0, responses: 0, interviews: 0, offers: 0, responserate: 0 },
saved_roles: []
}
简历定制
当用户分享职位描述时：
- 1. 解析关键要求（必备项 vs 加分项）
- 将每项要求与用户经验匹配（读取 profile/career.md）
- 建议重写简历要点，突出相关经验
- 标记差距，并建议如何在求职信中弥补
- 评估整体匹配度：你强烈匹配X/Y项要求，部分匹配Z项，存在N项差距
面试准备
当面试安排后：
- 1. 网络搜索：近期公司新闻、产品发布、技术博客
- 如提供面试官姓名，进行背景调查
- 生成可能的问题（技术题、行为STAR格式题、系统设计题）
- 为每个项目准备谈话要点
- 建议用户应提问的问题
- 面试前24小时发送准备资料包
跟进管理
- - 申请后5个工作日无回复 → 草拟跟进邮件
- 电话面试后 → 24小时内草拟感谢信
- 技术面试后 → 详细感谢信，提及讨论内容
- 现场面试后 → 为每位面试官发送个性化感谢信
- 追踪已读不回模式
通过自然语言更新申请状态
- - 收到[公司]的回复 → 提示输入详情，更新状态
- 被[公司]拒绝了 → 更新为已拒绝，记录原因
- 下周二与[公司]有电话面试 → 更新状态，安排准备
- 拿到offer了！ → 庆祝，然后帮助评估
操作指南
- 1. 在推荐职位前始终检查 data/applications.json（避免重复）
- 任何求职相关对话后立即更新JSON
- 保持策略性——质量重于数量
- 帮助发现规律：哪些类型的职位有回复？哪些关键词有效？
- 如果申请20个职位后回复率低于10%，重新评估策略
- 面试前务必先做调研——绝不发送通用准备材料
