Indeed HIDES the posted date even when you sort by date first which is a CRITICAL problem because you may end up wasting your time on a job that already has hundreads of applicants because the postings was weeks ago. However, you would never know, thus wasting your time even more.

LinkedIN does "Repostings" now where companies pay LinkedIN to have fake evergreen jobs that just stay up forever to collect your information with no intention of hiring anyone. This AND, fake gig platforms FLOOD the site so its hard to find real jobs. (Like Data Annotation).

I am just so tired of dealing with the fake job postings that I decided to take matters into my own hands. I figured I would try and figure out what companies work with what ATSs (Application Tracking Software) so I can just have a huge list of companies where I can just apply directly to them without Indeed or LinkedIN as the middle man.

Well, after about 2-3 weeks of collecting data from search engines, subdomain enumeration, etc - I was able to create an OpenSource ATS Aggregator job tracking app. It's called "OpenPostings" and it pulls in FRESH job postings that either:

1. Has a posted date that is within 24 hours (So you know the posting is fresh) 2. Has no posted date, and the job is still up/available.

This has over 31000+ companies where it is able to pull over 500000+ FRESH jobs DAILY.

I had Codex scaffold out a React Native application while I fixed all of its mistakes.

Here are all the ATSs it supports: • Workday • Ashby / ashbyhq • Greenhouse / greenhouse.io • Lever / lever.co • Jobvite / jobvite.com • Applicantpro / applicantpro.com • Applytojob / applytojob.com • Theapplicantmanager / theapplicantmanager.com • Icims / icims.com • Recruitee / recruitee.com • Ultipro / ukg • Taleo / taleo.net • BreezyHR • ApplicantAI • Career Plug • Career Puck • Fountain • Getro • HRM Direct • Talent Lyft • Talexio • Team Tailor • Talent Reef • Manatal • Zoho • BambooHR • Gem • Jobaps • Join • Saphrcloud

Once I had set it up and it was working for me to apply to a bunch of real jobs without Indeed or LinkedIN, I decided to setup an MCP server to it so you can have your AI/LLM model apply to jobs for you (As long as your AI has access to a browser). Here are the AI commands you can do with my app:

• Get back your applicant information • Find jobs that are related to your skillset • Generate dynamic cover letters based on your resume, experience and the job posting itself • Apply to jobs for your and update the app to show that an AI applied to a job for you in the application tracking section of the app (Requires your AI to have access to a browser)

- Its free - Its Open Source

It already has 39 stars on Github and people have said on Reddit and Youtube that it IMMEDIATELY helped them find REAL jobs. No recruiters, no sign up, no ads, nothing except a local seeded database, a react native app, and tons of jobs to apply to.

I genuinely hope that this gets ALOT of traction with the HOPE that we MAKE Indeed and LinkedIn ACTUALLY be more innovative rather than making it so hard for people to get jobs.

Comments URL: https://news.ycombinator.com/item?id=47807107

Points: 4

# Comments: 2

It persists the point-to-point protocol by using tor onion-style rendezvous for cross-network reconnects, but otherwise relies on the normal webRTC data channel for direct communication. (like PPP)

The result is that it lets apps talk directly (desktop ↔ mobile) without servers, accounts, or cloud infra.

I first thought of this when building an open source social media scheduler app where your content is on your desktop (or mobile) but you want to access the scheduler on your phone (or desktop). A server is an obvious choice but since this is open source, I wanted to avoid the need for users to setup a server or vpn, etc.

However this also has applications beyond just that: - Personal file sync (Dropbox alternative) - Password manager cross-device sync - Notes/tasks syncing (offline-first apps) - Home automation control (phone to local hub) - Game state sync (local-host multiplayer) - Camera/IoT monitoring without cloud - Clipboard sharing between devices - Local-first databases (CRDT peer sync) - Encrypted chat without central servers - Dev tools: logs/metrics between local machines

I personally tested this for windows and linux with the python bindings, but would like people to test this out on macOS if possible.

This supports: - Windows - Linux - macOS

Language options: - JavaScript/TypeScript - Python - C++ - PHP - Java - Rust

Why didn't I just use tor's onion relays fully for traffic if I am using it for rendezvous? Because tor's relays are optimized for low-bandwidth traffic and have high latency, which is not ideal for real-time communication. By using tor only for the initial rendezvous, I can establish a direct peer-to-peer connection using webRTC, which is much faster and more efficient for ongoing communication.

The protocol is still in early stages, but the core features are working. I’m looking for feedback on the design and implementation, as well as help with testing and documentation.

Comments URL: https://news.ycombinator.com/item?id=47522181

Points: 2

# Comments: 1

Chrome Extension: https://chromewebstore.google.com/detail/youtubetv-party-sha...

Github repo: https://github.com/Masterjx9/YoutubeTVParty