boxmoe_header_banner_img

Hello! 欢迎来到阿虪的开发世界~

加载中

文章导读

dossier:把求职装进自己服务器的自托管求职助手


avatar
阿虪 29 10 月, 2026 307
dossier

中文版

dossier(GitHub:https://github.com/AienTech/dossier ,2026-10-03 创建,今晨抓取时已 2,757★)是一个自托管的求职应用:一份个人档案,对应每一份工作,生成量身定制的简历和求职信,并把从起草到拿到 offer 的每一次投递都管起来。

用法是:先把职业经历、技能、项目、教育写进一份档案(也可以从现有简历 PDF / Word / Markdown 导入);粘贴一份招聘启事,它会拆出要求、算出你和岗位的匹配度、标出缺口,然后只用档案里的事实写一份定制简历和求职信——项目明确承诺“绝不编造经历”,档案里没有的东西只会显示为缺口,不会替你编上。改稿在聊天里完成,可导出 PDF;投递状态用看板管理,还有漏斗仪表盘。找工作可走 JSearch(LinkedIn、Indeed 等)或 Adzuna,也能看 YC 创业公司和 Hacker News 每月招聘帖;可选的每日求职 agent 只会准备草稿、绝不替你投递或发送。

为什么值得了解

  • 求职时最费时间的活就是“针对每个 JD 改一版简历”——这个项目把这条流水线做成了一个本地工具,而不是又一个要交简历数据的网站
  • 数据在自己的 PostgreSQL 里,模型可以选本地的 Ollama / LM Studio,也可以接 OpenAI 兼容、Anthropic 或 Mistral;Gmail / 日历、Claude 连接器都是可选的,不发送任何东西前都要你确认
  • “绝不编造经历”这条原则很关键:定制可以润色表达,但不能无中生有,这正是简历工具最该守住的底线

English Version

dossier (GitHub: https://github.com/AienTech/dossier , created 2026-10-03, 2,757 stars at this morning’s fetch) is a self-hosted job search app: one profile, a tailored resume and cover letter for every job, and every application tracked from draft to offer.

The flow is: put your career history, skills, projects and education into one profile (or import an existing resume as PDF, Word or Markdown); paste a job posting and it extracts the requirements, scores how you match, and flags the gaps — then writes a tailored resume and cover letter using only facts from your profile. The project promises it “never invents experience”: anything the posting asks for that your profile doesn’t have shows up as a gap, not as invented history. Revisions happen in a chat, documents export as PDF, applications are managed on a board, and a funnel dashboard shows progress. Job search runs through JSearch (LinkedIn, Indeed and more) or Adzuna, plus YC startups and Hacker News’ monthly hiring thread; an optional daily job agent only prepares drafts and never sends or applies on your behalf.

Why it’s worth knowing

  • The most time-consuming part of job hunting is “rewrite the resume for every JD” — this project turns that pipeline into a local tool instead of yet another site you hand your resume data to
  • Your data lives in your own PostgreSQL database, and the model can be a local one through Ollama / LM Studio, or a hosted OpenAI-compatible, Anthropic or Mistral model; Gmail / Calendar and Claude connectors are optional, and nothing is sent without your confirmation
  • The “never invents experience” principle matters: tailoring can polish wording, but it must not invent history — exactly the line a resume tool should hold


评论(0)

查看评论列表

暂无评论


发表评论

表情 颜文字

插入代码