Job Journal

Interview your career, customize your resume.

Job Journal is a local-first CLI tool that helps you build a rich corpus of your professional experience through conversational interviews, then uses that corpus to generate tailored resumes in your own voice. Every resume bullet is selected from your corpus -- never fabricated by AI.

Built for Claude Code as a first-class interface.

Quick Start

# Clone and install
git clone https://github.com/dontoisme/Job-Journal
cd Job-Journal
pip install -e .

# Initialize
jj init

# Start building your corpus (best with Claude Code)
jj interview

Optional Dependencies

pip install -e ".[web]"      # FastAPI web dashboard
pip install -e ".[gmail]"    # Gmail email tracking
pip install -e ".[dev]"      # Testing and linting

How It Works

1. Interview -- Job Journal walks through your career year-by-year, role-by-role, extracting achievements, skills, and stories in your own words.

2. Corpus -- Your experiences are stored in a searchable corpus (~/.job-journal/corpus.md) that you can edit and refine.

3. Discover -- Search job boards (Greenhouse, Lever, Ashby) and score opportunities against your corpus.

4. Apply -- Job Journal matches your corpus against the job description and generates a tailored resume. Every bullet is selected verbatim from your corpus.

5. Track -- Monitor application status, track email responses via Gmail, and view pipeline analytics.

With Claude Code

New here? Paste the install prompt in docs/getting-started-with-claude.md into Claude Code from the repo root — it installs the tool, sets up your profile, builds your corpus, and walks you through your first application with your own details.

Job Journal is designed around Claude Code slash commands as the primary interface:

Skill	Purpose
/interview	Conversational interview to build your corpus
/apply <url>	Full application workflow with fit scoring and resume generation
/jobs [query]	Search and score job opportunities
/greenhouse	Search Greenhouse job boards with saved auth

CLI Commands

Core

Command	Purpose
jj init	Initialize Job Journal
jj interview [role]	Start/continue building your corpus
jj import-base <file>	Import existing base.md file
jj stats	Show corpus and application statistics
jj serve	Start the web dashboard at localhost:8000

Corpus Management (jj corpus)

Command	Purpose
jj corpus sync	Sync entries from corpus.md to database
jj corpus list	List entries with filtering by tags, role, category
jj corpus search <query>	Full-text search across entries
jj corpus stats	Corpus statistics breakdown
jj corpus suggestions	View improvement suggestions from JD gap analysis
jj corpus edit	Open corpus.md in your editor

Resume Generation (jj resume, jj gdocs)

Command	Purpose
jj resume list	List all generated resumes
jj resume show <id>	Display a specific resume
jj resume validate	Validate resume bullets against corpus
jj gdocs generate <variant>	Generate tailored resume via Google Docs
jj gdocs setup	Configure Google Docs API

Job Discovery (jj greenhouse)

Command	Purpose
jj greenhouse setup <har>	Import auth from HAR file
jj greenhouse poll	Search for matching jobs
jj greenhouse config	Configure search defaults

Application Tracking (jj app, jj email)

Command	Purpose
jj app status	Show current application pipeline
jj app timeline	Application timeline over time
jj email setup	Configure Gmail API
jj email sync	Full sync: confirmations + updates + pairing
jj email verify	Check for application confirmation emails
jj email updates	Check for interview/rejection emails
jj email pair	Match emails to applications

Background Worker (jj worker)

Command	Purpose
jj worker start	Start background sync daemon
jj worker status	Show worker status
jj worker run-task <type>	Execute a specific task

Resume-JD Scoring

When applying, resumes are scored against the job description on a 100-point rubric:

Category	Weight
Summary alignment	25 pts
Skills coverage	25 pts
Bullet relevance	35 pts
Keyword density	15 pts

Target: 80+ before submitting.

Data Location

All data stays on your machine in ~/.job-journal/:

~/.job-journal/
├── profile.yaml      # Your contact info and preferences
├── config.yaml       # Settings and variant definitions
├── corpus.md         # Your professional corpus (human-readable, editable)
├── journal.db        # SQLite database (16 tables)
├── applications.csv  # Application tracking export
├── templates/        # Resume templates (docx)
└── resumes/          # Generated resumes

Integrations

Service	Purpose	Setup
Google Docs	Resume generation + PDF export	jj gdocs setup
Gmail	Application email tracking	jj email setup
Greenhouse	Job board search	jj greenhouse setup
Google Maps	Geographic company discovery	Config API key

Philosophy

• Your voice matters: Job Journal captures how you describe your work, not generic resume-speak.
• SELECT, don't COMPOSE: Every resume bullet comes verbatim from your corpus. No AI fabrication.
• Slow and thorough: Building a good corpus takes time. Each role gets a dedicated interview session.
• Local-first: Your career data stays on your machine in human-readable formats.
• AI-assisted, human-controlled: Claude helps extract and organize, but you approve everything.

Contributing

See CONTRIBUTING.md for development setup and guidelines.

License

MIT