What is a LinkedIn auto-apply bot?
A LinkedIn auto-apply bot is a tool that searches job postings and submits applications with less manual typing. Most are cloud services or scripts that hold your credentials and apply out of sight. With ego(lite), your own coding agent does the applying in a visible browser on your Mac, using the LinkedIn session you are already signed in to, and pauses before every submission.
How does the AI job application bot work in ego(lite)?
You give the agent your target roles, an answer bank, and a daily cap using the prompt on this page. The agent opens LinkedIn Jobs in its own Space, screens each matching posting, fills the application from your answers, and stops at the review step so you confirm it. At the end it returns a log of what was submitted, skipped, or left for your review.
Does it work with LinkedIn Easy Apply?
Yes. Easy Apply is the main flow this workflow automates: the agent steps through the form pages, attaches the resume you chose, and answers screening questions from your answer bank. A multi-step form is fine. A question your answers do not cover pauses the task instead of receiving a guess.
What about jobs that redirect to a company career site?
Postings that leave LinkedIn open in the same visible browser, so the agent can continue onto the external application form and fill it with the same approved answers. If the external site requires creating an account, uploading extra documents, or another decision you have not approved, the agent pauses and hands the Space to you.
Will LinkedIn restrict my account for auto-applying?
It can. LinkedIn does not publish an official application limit, but it throttles high-volume activity, and automated activity can conflict with its User Agreement. This page does not claim any volume is safe. Keep the daily cap small, review every application before it goes out, and stop immediately if LinkedIn shows a warning, CAPTCHA, or verification screen.
How many jobs can it apply to per day?
You set the cap yourself in the prompt, and the agent stops when it is reached. A small, targeted batch of roles you actually match tends to get more responses than mass submissions, and it keeps your activity closer to normal manual use. The workflow is designed for bounded, reviewed runs, not an unlimited application stream.
Do I upload my resume or LinkedIn password to a service?
No. ego(lite) runs on your Mac and the agent works through the browser context you authorize, so your existing LinkedIn session does the applying. The agent never asks for or types your password, and your resume goes only where you would upload it yourself: the application form.
Can it answer screening questions?
Only from the answer bank you wrote. Common questions about work authorization, notice period, salary expectations, and years of experience get your prepared answers. Anything outside the bank pauses the task so you answer it yourself. The agent is instructed never to invent an answer on your behalf.
Can it tailor a cover letter or short answers for each job?
You can ask your agent to draft a short cover note or answer from materials you provide, such as your resume and a summary of the role. The draft still sits in the review step until you approve it, and the workflow never fabricates experience, credentials, or authorization status to improve a match.
Is this LinkedIn auto-apply workflow free?
ego(lite) is free to download for Mac, and there is no per-application pricing. The agent doing the work is the coding agent you already run, such as Claude Code or Codex, so its token usage and any provider costs still apply.
How can I automate LinkedIn applications without getting banned?
There is no guaranteed safe volume or way to make automation invisible. Keep runs small and targeted, use the account and session you own, review every form, and stop at warnings, CAPTCHAs, verification, or rate limits. Do not rotate fingerprints, mimic mouse movements to evade detection, or bypass LinkedIn controls; automated activity can conflict with LinkedIn's User Agreement.
Can it tailor my resume for each LinkedIn job without inventing details?
Yes, it can reorder and rephrase facts you provide to match a role's terminology, while keeping your real employers, dates, skills, and outcomes. Ask for a plain-text ATS-friendly copy, require missing evidence to be flagged, and review every edit before upload. It must not add credentials, metrics, or responsibilities you cannot defend.
Can I search LinkedIn and other job boards in one workflow?
You can use the same agent workflow across LinkedIn, Indeed, and company career pages that your authorized browser session can open. Keep each site's account state and rules explicit, record the source URL, deduplicate by company and role, and pause when an external ATS asks for a new account or an unfamiliar document.
Can it monitor new LinkedIn jobs and send me alerts?
The agent can run a bounded search on a schedule, compare the results with a previous CSV or Markdown log, and prepare an alert or review list when new matches appear. A scheduled job should notify you rather than submit unattended, and it must stop if the session expires or LinkedIn presents a verification screen.
How do I filter for remote work, visa sponsorship, and real job links?
Put those constraints in the search brief and require a second fit check before opening an application. Verify the work location in the full posting, treat sponsorship language as an employer claim to confirm, follow redirects to the employer's real career domain, and mark a suspicious or duplicate URL as needs review instead of applying.
Can it message recruiters or track follow-up after applying?
It can draft a personalized follow-up and keep a source-linked log of dates, contacts, and responses, but sending a connection request or message should remain a separate human-approved step. Do not mass-message recruiters, scrape private contact data, or imply that automation guarantees a reply.
What happens when LinkedIn changes its page or form?
A visible agent can re-read the current page and often adapt to changed labels, but no automation is immune to a DOM or workflow change. Use semantic labels where available, verify each required field and attachment, keep a trace or screenshot of failures, and stop for review rather than relying on stale selectors or repeated retries.
Can I connect this LinkedIn job workflow to n8n?
Yes, n8n can orchestrate triggers, spreadsheets, and notifications around a browser task, while ego(lite) handles the visible, logged-in steps your agent is authorized to perform. Keep credentials in n8n's secret store, pass only approved fields, require a human gate before submission, and stop on CAPTCHA, rate limits, or unknown questions.