The profile files that power every downstream command live inDocumentation 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.
.claude/skills/job-application-assistant/ and in the repo-root CLAUDE.md. Running /setup populates them automatically from your documents folder, a pasted CV, or an onboarding interview — but you can edit any of them directly at any time. Changes take effect immediately on the next /apply, /scrape, or /rank run, with no reload step required.
Profile File Reference
CLAUDE.md (repo root)
The entry point for every Claude Code session in this repo. Contains your full candidate summary — identity, education, experience, skills, goals, and target sectors — alongside the workflow rules and the verification checklist that /apply enforces at the end of every application. When Claude reads the repo for the first time in a session, this is the first file it sees.
Edit this file when you want to update your top-level profile, change your target sectors, add a new position, or adjust your deal-breakers. It is the single source of truth for the high-level summary that informs fit scoring.
01-candidate-profile.md
The structured, data-dense version of your CV. Organized into clearly delineated sections:
- Identity — name, address, phone, email, LinkedIn, GitHub, languages, employment status, and commute constraints
- Education — degrees in tabular form with period, institution, and key topics
- Professional Experience — each role as a structured block with bullets for responsibilities and achievements
- Independent Projects — freelance, open-source, and personal work outside employment
- Technical Skills — programming languages with proficiency and frameworks, domain expertise, and software/tools
- Publications — peer-reviewed work with DOI links
- Awards — competitions, hackathons, and recognitions
- References — named contacts with title, company, and contact details
/apply drafter reads when building experience bullets and the skills section. The more concrete detail you put here — specific tools, measurable outcomes, named projects — the more precisely the output can be tailored per role.
02-behavioral-profile.md
Your personality and behavioral assessment results. Typically structured around a PI, DISC, Myers-Briggs, or self-assessment framework:
- Named behavioral traits with descriptions
- Identified strengths and growth areas
- Environment description — what conditions you thrive in (autonomy, structured process, fast-paced teams, etc.)
- What energizes you and what drains you professionally
/apply scoring step and the “personal fit” paragraph in cover letters.
03-writing-style.md
Rules for tone, structure, and voice applied whenever /apply writes prose. Includes explicit do’s and don’ts (for example: no em-dashes, avoid clichés, no passive constructions), a description of the preferred sentence rhythm, and patterns extracted from past applications that worked well. Edit this file when you want to enforce a specific style preference or correct a recurring pattern in generated output.
04-job-evaluation.md
The scoring framework used in /apply Step 1. Contains:
- Skill match areas — strong, moderate, and weak zones calibrated from your actual background
- Career goals — what you are optimizing for in your next role
- Motivation filters — sector preferences, role-type preferences, and what genuinely interests you
- Deal-breakers — hard constraints that veto a posting regardless of score (location, travel, industry, etc.)
- Calibration from past applications — notes on which role types have historically been good or poor fits
/apply and /rank run.
05-cv-templates.md
Instructions the CV drafter follows when building a tailored cv/main_<company>.tex. Contains:
- The moderncv banking-style document structure
- Profile statement templates organized by role type (for example: technical/ML roles vs. domain-specialist roles)
- Section ordering guidance for different role types
- Relevance-weighted cutting rules for staying within the 2-page hard limit
- ATS parseability requirements
06-cover-letter-templates.md
Instructions the cover letter drafter follows when building cover_letters/cover_<company>_<role>.tex. Contains:
- The
cover.clsdocument structure with named macros (\lettercontent,\closing,\signature) - Opening paragraph patterns and salutation rules
- Closing formulation options (including language-specific variants such as Danish “Med venlig hilsen”)
- Hard 1-page limit and word budget guidance (250–300 words of body text)
- Known template pitfalls (for example: the
\lettercontent{}+\begin{itemize}interaction that breaks compilation)
07-interview-prep.md
STAR-format examples drawn from your actual experience, organized for reuse in interviews. Also contains:
- A question bank of likely interview questions for your target role types
- Talking points for each question referencing real projects and outcomes
- Questions the candidate should ask the interviewer
- A roleplay framework for practicing with Claude
Which Files to Edit for Common Tasks
| File | What to change |
|---|---|
CLAUDE.md | Full profile (name, education, experience, skills, goals) |
01-candidate-profile.md | Structured version of your CV data |
02-behavioral-profile.md | Your behavioral assessment or self-assessment |
04-job-evaluation.md | Skill match areas, career goals, motivation filters |
05-cv-templates.md | Profile statement templates for different role types |
07-interview-prep.md | STAR examples from actual experience |
search-queries.md | Job search queries for your skills and location |
Profile Depth Tips
The single biggest factor in output quality is how much detail your profile contains. A thin profile produces generic applications; a detailed one enables genuinely tailored output. Skills in context beat bare skill lists. Compare:| Thin | Rich |
|---|---|
Python, machine learning | Built ML pipelines for customer churn prediction in Python (scikit-learn, pandas); deployed model as a REST API serving 40k daily predictions |
Project management | Led a cross-functional team of 6 to deliver a data platform migration on time; managed stakeholder reporting and sprint planning in Jira |
Communication | Presented quarterly business insights to C-level stakeholders; translated model outputs into executive dashboards in Power BI |
/apply specific tools, named outcomes, and concrete contexts to reframe per role. Vague entries like “Python” or “good communicator” get reused as-is; rich entries get selectively emphasized, cut, or reframed based on each posting’s keywords.
Role descriptions matter as much as titles. List what you actually did — specific projects, tools used, responsibilities, and measurable achievements — not just the job title and company name. The system uses these descriptions to match your experience against posting requirements, so an entry that reads “Responsible for data analysis” gives it far less to work with than one that names the methods, tools, and outcomes.
All onboarding paths benefit from the same principle. Whether you run /setup from your documents/ folder, paste a single CV, or walk through the interview, richer input at every step produces sharper, more accurately targeted output.