linkfetch

A local tool that scrapes your own LinkedIn profile and emits tailor-compatible YAML, so you can bootstrap or refresh your tailor profile from LinkedIn instead of typing everything by hand.

linkfetch is not a website and does not host anything. It runs on your machine and drives a real Chromium browser that you log into. It captures only the authenticated user's own profile.

How it works (two phases)

1. Capture (browser, runs rarely): log in once in a persistent Chromium profile, then visit each section's LinkedIn details page (/in/<you>/details/experience/, /details/education/, …), auto-expand all "…see more" / "Show all" controls, and save the rendered HTML to disk.
2. Parse (pure, no browser): read the saved HTML, extract structured data with lxml, and write YAML files. Re-runnable offline; if a selector breaks you fix the parser and re-parse without touching LinkedIn again.

The scrape is deterministic — no AI. Fields come straight from the page. You enrich tags / position framings later in tailor.

Output

Sections that map onto tailor's schema are written as tailor-ready files (basics.yaml, experiences.yaml, skills.yaml, projects.yaml, certifications.yaml, awards.yaml). The rest get their own local-schema files (education.yaml, volunteering.yaml, recommendations.yaml, patents.yaml, courses.yaml, languages.yaml).

Review the output in data/output/, then copy the mapped files into tailor/data/profile/ and run tailor profile validate.

Requirements

• Python ≥ 3.10
• Playwright for capture (uv pip install playwright && playwright install chromium). Parsing works without it.

Install

cd ~/personal/prosperis/linkfetch
uv venv --python 3.10
uv pip install -e ".[dev,browser]"
playwright install chromium

Quick start

uv run linkfetch doctor                  # check config / Playwright / session
uv run linkfetch login                   # opens browser; log in to LinkedIn once
uv run linkfetch capture --vanity your-name   # save section HTML to data/captures
uv run linkfetch parse                   # HTML -> YAML in data/output
# or do both:
uv run linkfetch run --vanity your-name

--vanity is the slug in your profile URL: linkedin.com/in/<vanity>.

Legal note

LinkedIn's Terms of Service restrict automated access. This tool is for personal use on your own profile: it uses a real logged-in browser at a human-ish pace, reads only your own data, and stores nothing remotely. Use it on your own account and at your own discretion.