jobjob 2.0.0

Generate tailored resumes and cover letters at scale.

jobjob

Generate tailored resumes and cover letters at scale. Given a job-description PDF, jobjob runs five sequential Claude calls (sharing one cached context prefix) to parse the role, select credential highlights, generate a cover letter, and produce a skills gap analysis — writing local artifacts and (optionally) uploading to Google Drive.

New here? If you have never used a terminal before, start with the Quick-Start Guide — it walks through every step.

Install

pdm install            # or: pip install -e .

Requires Python ≥ 3.12.

Configuration

Configuration is split into two disjoint tiers (no overlap, validated at load):

• App config — config/.env, machine-local and gitignored (only config/.env.template is committed). One jobjob instance: secrets, local paths, output IDs, and the profile registry.
• Profile config — <profile-repo>/config/.profile, committed inside each profile repo. The active profile's applicant identity and resume template.

cp config/.env.template config/.env
# then edit config/.env with your values (incl. a JOBJOB_PROFILE_* registry entry)

A profile is a separate git repo of content jobjob tailors from (content/, reference/, config/.profile, optional prompt/). The app config registers profiles by path and selects the active one; switch profiles to swap the whole content set (e.g. one content set per role type). See docs/setup.md for the full reference.

Variable	Tier	Purpose	Required
ANTHROPIC_API_KEY	app	Anthropic API key	Yes
GOOGLE_CREDENTIALS_FILE / GOOGLE_TOKEN_FILE	app	Google OAuth	Drive/Sheets only
APPLICATIONS_FOLDER_ID	app	Applications-root Drive folder id	No
APPLICATIONS_LOCAL_DIR	app	Synced Drive mirror (local)	No
LINKEDIN_SHEET_ID	app	Contacts spreadsheet id (enrich)	enrich only
JOBJOB_PROFILE_<NAME> / JOBJOB_ACTIVE_PROFILE	app	Profile registry + active	Yes
RESUME_TEMPLATE_ID	profile	Resume-template Google Doc id	Drive only
APPLICANT_NAME / _PHONE / _EMAIL / _LINKEDIN	profile	Cover-letter header	No

Config is read once at the entry point and injected into the workflow; feature modules take plain parameters and never read the environment themselves.

For obtaining API keys and Google OAuth credentials, see docs/credentials-setup.md.

Usage

# Generate a resume + cover letter for a job-description PDF:
jobjob apply path/to/job_description.pdf          # full run (Drive + local)
jobjob apply path/to/job_description.pdf --skip-drive -o ./out  # local only

# Import a LinkedIn profile screenshot into the contacts sheet:
jobjob enrich path/to/profile.pdf
jobjob enrich path/to/profile.pdf --dry-run       # parse only, no sheet write

See docs/usage.md for the full flag reference and examples.

Local output (-o <dir>, default ./out/<date> - <Company> - <Role>/): cover letter (.pdf/.docx), skills_analysis.json, a copy of the JD, the resume PDF, and summary.json.

Google Drive folder (YYYY-MM-DD - Company - Role) holds exactly four artifacts: the README (Google Doc), the JD PDF, the customized resume (Google Doc), and the cover letter (Google Doc).

Resume templates

The active profile's content/templates.toml defines resume variations by archetype (e.g. leadership_biotech, leadership_research), each with a Google Doc doc_id and keywords/description. The workflow uses --template <name> if given, otherwise asks the model to classify the JD against the archetypes. With no templates.toml, a single RESUME_TEMPLATE_ID is used as the default. (The repo's static/content ships defaults used when no profile is active — e.g. tests and a fresh clone.) See docs/usage.md for the --template flag.

Architecture

• jobjob/config.py — central Settings, loaded from config/.env + the active profile's config/.profile; jobjob/loader/profiles.py resolves the registry.
• jobjob/ailib/ — query (retry + cache orchestration), client/ (thin provider adapters; anthropic.py), session.py (AIClient: cached context + token usage), cachemanager.
• jobjob/loader/ — auth, prompt/content/reference loaders, static-file location.
• jobjob/structure/ — typed domain models (job description, highlight, skill, …).
• jobjob/apply/generate/ — the AI steps (parse, highlights, resume, cover letter, skills).
• jobjob/apply/output/ — cover-letter PDF/DOCX builders.
• jobjob/gapi/ — Google Drive/Docs I/O.
• jobjob/apply/workflow.py — orchestration; jobjob/__main__.py — CLI.

Adding another AI provider is a single new adapter under jobjob/ailib/client/.

Webapp

A local dashboard for managing the queue, launching jobs, and reviewing results is in webapp/. See webapp/README.md for setup and usage.

Development

pdm run python -m unittest discover -t ./ -s tests/jobjob   # or: pytest