Skip to main content

Documentation Index

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

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

/setup is the first command you run after forking the repository. It reads your career documents, asks the right questions, and fills in every profile and configuration file that the rest of the framework depends on, so /scrape knows what to look for and /apply knows how to write for you. You can re-run it safely as your materials evolve, and you can target individual sections with --section when you only want to update one part of your profile without redoing everything else.

Syntax

/setup
/setup --section <name>
The --section flag accepts a section name from Path C (for example, search to update your job search configuration). Only search is called out in the command’s built-in guidance as a specific shortcut.

Choosing a path

When you run /setup without arguments, it scans the documents/ folder and presents three paths. The right choice depends on what you already have.
Best when: you have a CV, LinkedIn export, diploma transcripts, reference letters, or past application files in documents/./setup reads every file in documents/cv/, documents/linkedin/, documents/diplomas/, documents/references/, and documents/applications/, cross-references them for date and title consistency, and merges extracted data into the seven profile skill files. Conflicts (for example, a role title that differs between your CV and LinkedIn export) are surfaced one at a time so you choose which version to keep.Path A is read-before-write and idempotent: re-running it after you add more documents will only propose genuinely new content, never duplicate or overwrite what is already there.
Drop your documents into documents/ before running /setup. See documents/README.md for the expected folder layout.

Section update mode

Use --section to update a specific part of your profile without re-running the full onboarding flow.

Files populated by /setup

After all three paths converge, /setup generates or updates the following files.
FileWhat it contains
CLAUDE.mdFull candidate profile and workflow rules, with [PLACEHOLDER] tokens replaced by your data
.claude/skills/job-application-assistant/01-candidate-profile.mdIdentity, education, experience, skills, publications, references
.claude/skills/job-application-assistant/02-behavioral-profile.mdBehavioral assessment (PI, DISC, or synthesized from interview answers)
.claude/skills/job-application-assistant/04-job-evaluation.mdPersonalized scoring framework with strong/moderate/weak match areas and career goals
.claude/skills/job-application-assistant/05-cv-templates.mdRole-specific profile statement templates
.claude/skills/job-application-assistant/07-interview-prep.mdSTAR examples from your actual experience (or stubs for Path A)
cv/main_example.texLaTeX CV with your real name, contact info, and most recent entries
.claude/skills/job-scraper/search-queries.mdTargeted search queries for /scrape, built from your role titles, skills, and location
Files 03-writing-style.md and 06-cover-letter-templates.md are framework files. /setup does not overwrite them. They contain writing rules and template structures that apply to all candidates and are only extended when Path A finds consistent patterns in your past cover letters.

Path A: cross-reference and merge behavior

Path A reads your existing skill files before extracting anything, so it knows what is already there. It then builds two buckets:
  • Additive changes: entirely new content not yet in any skill file. All proposed additions are shown in a grouped list before anything is written. You can approve all at once or skip individual items.
  • Conflicting changes: content that touches an existing entry but disagrees (a different date range for a job, a different graduation year). Conflicts are resolved one at a time.
Behavioral inferences drawn from your LinkedIn About section or reference letters are always labeled as inferred, so you can review them before relying on them in applications.

Job search configuration

Regardless of which path you use, /setup ends with a set of questions about your job search (Section 9 in Path C, or follow-up questions after the document merge in Path A):
  • Role titles to search for (3-8 specific titles)
  • Key skills as search terms (3-5 distinctive, searchable skills)
  • Target companies (optional)
  • Geographic scope and commute limits
  • Which job portals to use
Based on your profile, /setup also proactively suggests role types you may not have considered, for example adjacent titles in your domain or leadership roles that fit your experience level.
Run /setup --section search later to update these queries as your priorities change, without touching any other part of your profile.

Next steps

After setup completes, the confirmation message tells you exactly what was written. The suggested follow-ups are:

/scrape

Search the configured portals for matching jobs right now

/apply

Run the full application workflow on any job posting URL

/expand

Enrich your profile from GitHub, portfolio sites, and course syllabi

/reset

Start fresh if you need to re-onboard from scratch

Build docs developers (and LLMs) love