AI Job Search
A Claude Code workflow where the agent searches job postings, scores fit against your profile, drafts a CV and cover letter, and helps prepare for interviews
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- The user's personal data (CV, salary expectations) is saved into the fork's files
- The agent reads third-party job posting content from job portals
Install
Manual install
gh repo fork MadsLorentzen/ai-job-search --clone
cd ai-job-searchThen inside Claude Code run /setup, /scrape and /apply.
This is third-party code. Review the repository files before installing.
What it does
The project is forked to your own account, then /setup builds a user profile from a documents folder, a pasted CV, or a short interview with the agent. /scrape searches several job portals, dedupes results and sorts them by fit, while /rank batch-scores a larger list against a shared criteria framework. /apply runs a two-agent chain: one drafts a LaTeX CV and a cover letter for the specific posting, the other reviews the result and asks for revisions. The core workflow is not tied to a country, while the job-portal search skills themselves are built for the Danish labor market.
Who it is for. For job seekers who want to automate posting search, CV drafting and cover letters through Claude Code.
Good fit when
- You need a quick fit assessment for a job posting against your profile
- You need a CV and cover letter tailored to a specific posting
- You have access to Danish job portals, or are ready to adapt the search skills to your own
Not a fit when
- The ready-made job-search skills target Denmark; other markets need custom rewrites
- You are not comfortable putting personal data (resume, salary expectations) into your own fork of the repo
- You don't have a LaTeX distribution to build the CV and letter PDFs
Example request
Run /scrape to find postings matching my profile, then prepare an application for the best one via /applyLimitations
Requires a Claude Pro/Max/Team subscription or API token billing, plus a LaTeX distribution to build the CV and letter PDFs. A fork of a public repo is always public on GitHub; for private data the author recommends keeping a private repository with this project as upstream instead. It sends job posting content to the agent, but per the author it does not follow instructions embedded in the posting text.
How to disable. Delete your fork of the repository or the local copy; nothing else is left on the system.
Security check
- The user's personal data (CV, salary expectations) is saved into the fork's files
- The agent reads third-party job posting content from job portals
README in short
The README describes the project as a job-search automation framework on top of Claude Code, built by the author for their own search after a layoff. The core loop is /setup for the profile, /scrape for finding postings, and /apply for drafting a CV and letter with a second-agent review. Over ten extra commands exist, including /interview for interview prep and /rank for batch-scoring postings. The job-search skills target specific Danish job portals and are packaged as separate Bun CLI tools inside .agents/skills.
FAQ
Does it work with agents other than Claude Code?
The core workflow is stated to be language- and country-agnostic, and for other agents like Codex, Antigravity or Gemini CLI the README points to the AGENTS.md file and community forks.
Is LaTeX required?
Yes, the CV compiles via lualatex and the cover letter via xelatex; a minimal distribution like TinyTeX works with the extra packages installed.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI