career-tailor 0.1.0

Agent-assisted resume tailoring for job seekers.

Career Tailor

Career Tailor is a local-first command line tool that turns a master resume into job-specific drafts. It extracts a structured profile from your resume, compares that profile against a job description, and generates tailored Markdown, JSON, CSS, and optionally PDF output.

The goal is not to fabricate experience. The agent is prompted to preserve the facts in your profile and reframe them in the language of the job description.

Why

Many applicant tracking systems and recruiter workflows reward resumes that use the same language as the job posting. Career Tailor helps job seekers build truthful, role-specific versions of a resume without manually rewriting the same material for every application.

Install

uvx career-tailor --help

For local development:

git clone https://github.com/chriestensonb/career-tailor.git
cd career-tailor
uv sync
cp .env.example .env

You can also run the latest unreleased GitHub version directly:

uvx --from git+https://github.com/chriestensonb/career-tailor career-tailor --help

Set your model provider key in .env. By default the project uses a Pydantic AI Anthropic model string, but OpenAI model strings are supported too:

CAREER_TAILOR_MODEL=anthropic:claude-sonnet-4-6
ANTHROPIC_API_KEY=...

Or:

CAREER_TAILOR_MODEL=openai:<model-name>
OPENAI_API_KEY=...

Usage

Create a structured profile from an existing resume:

career-tailor ingest ./my-resume.md

Generate a tailored resume from a job description:

career-tailor tailor ./job-description.md

Skip PDF rendering when you only want Markdown and JSON:

career-tailor tailor ./job-description.md --no-pdf

The default profile path is data/profile.json. Generated outputs go to tailored/. Both directories are ignored by Git because they commonly contain personal information.

Data And Privacy

Career Tailor runs locally, but it sends resume and job description text to the model provider you configure. Do not use it with sensitive data unless you are comfortable with that provider's data handling terms.

The repository ignores:

• .env and other local env files
• data/, which stores extracted profile data
• tailored/, which stores generated resumes
• local caches and virtual environments

Development

uv sync --dev
uv run pytest
uv run ruff check .

If PDF rendering fails the first time, install Playwright's Chromium browser:

uv run playwright install chromium

Project Status

This is early alpha software. It is useful as a local CLI today, and the next good open-source milestones are provider examples, safer redaction workflows, more resume import formats, and package publishing.