Skip to content

Repository files navigation

JobHuntBot

Turn your AI coding agent into a persistent job-search system — from industry research and benchmark JDs to evidence-based resumes, live job matching, and application tracking.

把 Codex / Claude Code 从“帮我改一次简历”,变成一个持续运行的完整求职 Agent Harness。

What is JobHuntBot?

JobHuntBot is a job-search Agent Harness designed for AI coding agents.

It turns the job search process into a structured workflow:

  1. Analyze target role requirements
  2. Research industries and companies
  3. Build company trees (head / growth / small-but-high-quality)
  4. Extract capability models from benchmark JDs
  5. Mine real experience evidence
  6. Generate tailored resumes
  7. Verify live jobs and track applications

The core idea is simple: resume is not written first. It is derived from market requirements and your proven evidence.

How to Use

Quick Start

Clone the repository:

git clone https://github.com/5JjiaLin/JobHuntBot.git
cd JobHuntBot

Then load this project into Codex / Claude Code and copy the following prompt:

使用这个项目作为我的求职 Agent Harness:

https://github.com/5JjiaLin/JobHuntBot

请先读取项目中的 AGENTS.md 和 SKILL.md,理解完整工作流程。

然后严格按照项目定义的 Phase 流程引导我完成求职。

不要跳过阶段,不要直接生成简历。
从 Phase 1 开始,先询问我的目标岗位。

The agent should first load the workflow definition, then execute each Phase step by step.

Workflow

Phase Process Output
Phase 1 Role → Industry → Company Tree → Benchmark JDs → Capability Model 企业树、行业岗位核心能力
Phase 2 Real experience discovery and verification 个人经历
Phase 3 Evidence matrix → STAR → Resume generation → Truth audit 证据矩阵、简历
Phase 4 Recruitment track → Live job verification → Application priority 投递清单
Phase 5 Workspace initialization → Dashboard 求职 Dashboard

Core Features

Industry-aware Role Intelligence

先确定岗位所在行业,再研究该行业企业和岗位要求,而不是从一份 JD 开始。

Company Tree

企业按照目标行业和岗位划分:

  • Head / benchmark companies
  • Growth / mid-size companies
  • Small-but-high-quality companies

Small-but-high-quality Discovery

通过:

  • 官方资质认证
  • 行业榜单
  • 创投/产业数据库
  • 业务真实性验证

寻找低知名度但具备竞争力的企业。

Evidence-based Resume

能力模型 → 真实经历 → 证据矩阵 → STAR → 简历。

避免 AI 编造经历,只优化真实能力表达。

Local Workspace Dashboard

所有求职数据保存在本地 workspace,支持岗位跟踪和投递管理。

Project Structure

JobHuntBot/
├── AGENTS.md
├── SKILL.md
├── references/
├── assets/
├── dashboard/
├── templates/
├── scripts/
└── workspace/

Privacy

  • workspace 默认不上传 GitHub。
  • 不会自动扫描用户个人文件。
  • 个人经历只在用户授权后读取。
  • 不绕过登录、验证码、权限限制。

License

MIT

About

AI-native, local-first job search system: JD research → evidence-based resumes → live job matching → CSV-backed application dashboard.

Topics

Resources

Contributing

Security policy

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages