Autoapplier

Two tools in one repo:

1. Job Match Ranker -- a web dashboard that scores and ranks job postings against your resume
2. Auto-Applier -- a CLI bot that fills out and submits job applications automatically

---

Job Match Ranker

A local web dashboard at http://localhost:5050 that helps you figure out which jobs are worth applying to.

• Pulls fresh postings from SimplifyJobs
• Scrapes each job description
• Scores fit (0-100) using AWS Bedrock Claude based on experience match, recency, and location
• Ranked table you can filter by score, age, category, and applied status
• Track what you've applied to with one-click "Mark Applied"

python -m openclaw.web_ui
# Open http://localhost:5050

Click Score New Jobs to fetch and rank the latest postings. Click Score Unscored to process any jobs already in the ledger that haven't been scored yet.

---

Auto-Applier

A headless browser bot that opens a job posting, fills every field, uploads your resume, answers custom questions with an LLM, and optionally submits.

Supports Lever, Greenhouse, Ashby, Workday, and a generic fallback for other sites.

Apply to a single job:

python -m openclaw.applier "https://jobs.lever.co/company/job-id" \
  --company "Acme Corp" \
  --role "Software Engineer Intern" \
  --dry-run

Batch apply from SimplifyJobs:

python -m openclaw.applier --source simplify \
  --category "software engineering" \
  --max-jobs 10 \
  --dry-run

Apply to your top-ranked jobs from the scorer:

python -m openclaw.applier --source simplify --apply-top-scored \
  --min-score 75 --max-jobs 5 --dry-run

Drop --dry-run to actually submit.

---

Setup

pip install -r requirements.txt
playwright install chromium

Create a real_memory/ folder (gitignored) with your personal data:

real_memory/
├── profile.json   # Your info + answer bank (copy from profile.example.json)
├── resume.json    # Structured resume data
└── resume.pdf     # Your actual resume

The scorer and question answerer call AWS Bedrock Claude. Set up ~/.aws/credentials or the standard AWS_ACCESS_KEY_ID / AWS_SECRET_ACCESS_KEY env vars.

---

CLI Flags

Flag	What it does
--dry-run	Fill forms but don't submit
--force	Auto-submit even for top-tier companies
--human-in-loop	Pause for CAPTCHAs, auth walls, or missing fields
--headful	Show the browser window
--keep-open	Leave browser open after the run for manual review
--quality	Generate cover letters, richer answers (slower)
--score-jobs	Score new jobs from source (--source simplify required)
--list-scored-jobs	Print scored jobs from the ledger
--ledger-stats	Print ledger stats
-v	Verbose logging

Answer Bank

Add a questionBank array to profile.json to give consistent answers to common form fields (LinkedIn URL, "how did you hear about us", etc.). Matching is best-match-wins against form labels. Use __HUMAN__ for questions you want to answer yourself. See profile.example.json for the full template.

Privacy

real_memory/, test_memory/, .env, and all credential files are gitignored. Only source code, config templates, and this README get committed.