LIN · AI JOB-SEARCH AGENT

From job posting to submit-ready application — automatically.

Lin scans the market, scores every role against your real experience, writes two competing ATS-tailored résumés per job, and packages everything for you to review and send. While you sleep.

🎭 Explore the live demo View on GitHub

One rule, never broken: you always submit — Lin never applies on your behalf.

How it works

Six stages run on a schedule, twice a day — or any of them on demand, with one click.

1 · Scan

🔍 Find the roles

Sweeps career portals you track (plus optional LinkedIn, Indeed and Gmail job-alert channels), or just paste any posting URL.

2 · Score

🧮 Grade the fit

Every role gets a structured 0–5 evaluation — skills match, level strategy, compensation signals, posting legitimacy — plus a location-eligibility check.

3 · Stage

🎯 Verify it's real

Before spending effort, Lin confirms the posting is still live and actually accepting applications. Dead and stale listings never waste your time.

4 · Build

🛠️ Write the résumés

Two engines each craft a tailored résumé from your master profile — never inventing a word of experience. A quality gate checks page fill and length before anything ships.

5 · Finalize

📦 Package it

An automated ATS comparison picks the winning résumé, application answers are drafted, and a recruiter-named PDF lands in a tidy folder with a submit checklist.

6 · You apply

✅ Stay in control

You review the package and submit it yourself. Lin records it, then watches your inbox for rejections, interview invites and offers — updating the tracker automatically.

Two engines. One winner. Every time.

Résumé advice always splits into two camps — so Lin builds both and lets the evidence decide.

For humans

FORGE

Restructures your story per role: reorders achievements, rewrites the summary, leads with what this hiring manager cares about. A résumé that reads like it was written for the job — because it was.

For ATS robots

PATHFINDER

Surgically injects the job description's exact keywords while keeping your structure intact, so automated screeners score you the way they should.

Every role gets both → an automated comparison scores keyword coverage, structure and narrative → the winner ships. A rolling scoreboard tracks which engine wins more over time.
Both engines evolved from excellent MIT-licensed open-source projects — full credits in the repo.

An AI agent runs the whole thing

Lin is built on Hermes, an open agent platform — engineered the way an agent system should be:

Right model, right job. Cheap, fast models handle scanning, scoring and bookkeeping; a frontier model is reserved for the one task that deserves it — writing your résumé.
Deterministic where it matters. Tracking, packaging and data integrity run as plain code with 60 automated tests — no model in the loop, no surprises.
Everything on a dashboard. A funnel-rail view of your entire pipeline: filter, sort, bulk-act, flag a role for building, or run the whole chain with one ⚡ click.
Yours to inspect. The full system — skills, scripts, engines, dashboard — is open source under MIT.