Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/MadsLorentzen/ai-job-search/llms.txt

Use this file to discover all available pages before exploring further.

This guide covers every prerequisite for AI Job Search in full detail. The Quickstart gets you running in 15 minutes on the happy path; come here when you need exact install commands for your platform, want to understand why each dependency is required, or are troubleshooting a specific tool.

​
Prerequisites

​
Claude Code

Claude Code is Anthropic’s CLI that powers the entire framework. Install it globally with npm: 
npm install -g @anthropic-ai/claude-code
You need either an Anthropic API key or a Claude Pro/Team subscription to use Claude Code. See the Claude Code documentation for authentication details.

​
Python 3.10+

Python is required for the salary lookup tool. Check whether you have a compatible version: 
python --version
On Windows, py --version is often the most reliable check if python is not on your PATH.

​
Bun

The job portal CLIs are written in TypeScript and run with Bun. Install it for your platform: 
curl -fsSL https://bun.sh/install | bash
On Windows you can also use winget: winget install Oven-sh.Bun.

​
LaTeX

The /apply command compiles CVs with lualatex and cover letters with xelatex. Install a full LaTeX distribution for your platform:
Install MiKTeX. MiKTeX supports on-the-fly package installation, so missing packages (such as moderncv) are downloaded automatically on first compile.
Use lualatex for the CV — pdflatex frequently fails on modern MiKTeX installs with fontawesome5 font-expansion errors. lualatex handles the same sources cleanly.
The CV uses lualatex because pdflatex often fails on modern MiKTeX installs with fontawesome5 font-expansion errors. The cover letter uses xelatex because cover.cls requires fontspec for its custom Lato/Raleway fonts.

​
pdftotext (optional)

/apply runs an ATS parseability check on the compiled CV: it extracts the PDF’s text layer and verifies contact details, reading order, and keyword coverage the way an applicant-tracking system sees them. This requires pdftotext from poppler, which is separate from any LaTeX distribution: 
brew install poppler
If pdftotext is missing, /apply skips the mechanical check with a warning and falls back to a visual keyword review — everything else in the workflow runs normally.

​
Fork and Clone

Fork the repository to your GitHub account and clone it locally: 
gh repo fork MadsLorentzen/ai-job-search --clone
cd ai-job-search
If you prefer the manual route, fork on GitHub first, then clone your fork with git clone.

​
Install Job Search CLI Tools

Run the following from the repository root to install dependencies for all five job portal CLI tools: 
for tool in jobbank-search jobdanmark-search jobindex-search jobnet-search linkedin-search; do
  cd .agents/skills/$tool/cli && bun install && cd ../../../..
done
The linkedin-search install is optional: it has zero runtime dependencies and runs with plain bun; bun install only pulls TypeScript dev types. linkedin-search is country-agnostic and works for any location via the -l flag (e.g. -l "Berlin, Germany").
The four Danish portals (Jobbank, Jobdanmark, Jobindex, Jobnet) are built for the Danish market. If you are outside Denmark, use /add-portal to generate an equivalent search skill for your local job board — the command investigates the portal, scaffolds the CLI, and test-runs a live query before registering anything.

​
Run /setup

Start Claude Code from the repository root, then run the onboarding command: 
claude

/setup
Claude will detect what is in your documents/ folder and offer three onboarding paths:
  • Path A (documents folder): Drop your CV PDF, LinkedIn export, diplomas, reference letters, or past applications into documents/. Claude reads and cross-references everything before proposing profile updates. This path is idempotent — re-running it as you add more files will not duplicate or overwrite existing content; conflicts are surfaced for explicit resolution.
  • Path B (single CV import): Mention a file with @ or paste your CV text directly in chat. Claude extracts the data and asks follow-up questions for anything missing.
  • Path C (interview mode): Walk through structured questions section by section. Good for starting from scratch without existing documents.
All three paths converge on the same result: a complete set of populated profile files.

​
What Gets Populated

FileContent
CLAUDE.mdYour full candidate profile
01-candidate-profile.mdStructured education, experience, and skills
02-behavioral-profile.mdBehavioral assessment
04-job-evaluation.mdPersonalized skill match areas and career goals
05-cv-templates.mdProfile statement templates for your background
07-interview-prep.mdSTAR examples from your experience
cv/main_example.texYour LaTeX CV with actual details
search-queries.mdJob search queries for /scrape

​
Re-running Setup for a Single Section

You can update specific sections without re-running the full profile interview: 
/setup --section skills
/setup --section experience
/setup --section search
The --section search option is especially useful as your job search priorities evolve. It re-runs the search configuration interview and suggests role types you may not have considered based on your full profile.

​
Optional — Salary Benchmarking

The framework includes a salary benchmarking tool that works with any salary data you provide (union statistics, Glassdoor exports, survey data, or personal research). If you have data to add:
  • Option A: Create salary_data.json manually in the repository root. See tools/README_SALARY_TOOL.md for the expected format.
  • Option B: Convert from an Excel file:
pip install openpyxl
python tools/convert_salary_excel.py path/to/salary-data.xlsx --source "My Salary Data 2025"
This creates salary_data.json, which the /apply workflow uses for salary benchmarking. If you skip this step, the salary lookup is simply omitted. See Salary Benchmarking for more details.

​
Test the Workflow

With setup complete, test the full application workflow against a real job posting: 
/apply https://jobindex.dk/job/1234567
Or paste a job description directly: 
/apply [paste job posting text here]
Claude will evaluate fit against your profile, ask whether to proceed, draft a tailored CV and cover letter, have a reviewer agent critique the drafts, revise and compile both PDFs, and present the final output with a verification checklist.

Build docs developers (and LLMs) love

Get started for freeTalk to us