Resume Tailor — AI skill for Claude Code

resume-tailor

Turn a job description into a **fully tailored resume + cover letter**, in **DOCX and PDF**, with a **skills-gap report and a 0–100 match score** — while keeping your **clients, employers, locations and work durations locked** so they can never drift.

Every run rewrites the professional summary, the ordering of your skills, and the project context + bullets of *every* engagement so the technologies and responsibilities the job asks for sit front and centre — then force-reverts the client names and start/end dates from your master profile.

• CLI + local web app with an application tracker (every run saved, status, notes, match/coverage view, re-render).
• Any AI provider — Claude Code subscription, Anthropic / OpenAI / Gemini / OpenRouter API, or a local model (Ollama, LM Studio, llama.cpp, …). Pick it on the Settings tab.
• Multiple master profiles (e.g. Backend vs Data vs Platform); "Auto" routes a JD to the best-fitting one by required years + role.
• Optional one-box Proxmox LXC + Tailscale self-host — `deploy/DEPLOY.md`.

Quickstart

pip install -e .                 # from a clone;  or: ./setup.sh
resume-tailor init               # scaffolds profile/ + settings.yaml

# try it right now, no API key, no network:
resume-tailor tailor --engine mock --jd examples/sample_job.txt

$EDITOR profile/profiles/default/master_profile.yaml   # your real resume, once
resume-tailor web                # http://127.0.0.1:8000  → Settings tab picks the provider

Python 3.11+. LibreOffice is optional (pixel-faithful PDF from the DOCX); without it, a built-in renderer mirrors the layout.

With Docker

docker compose up --build        # http://127.0.0.1:8000 , data in ./rt-data

The container can't use the `