Skip to content
HeyJared

Open Source Spotlight

πŸš€ Welcome to Open Source Spotlight! I explore the best new open source GitHub repositories and tools - so you don't have to.

YouTube Β· IN Β· Official site

Indexed videos, last 90 days
11
Latest publication
Jul 25, 2026
Audience
~272 subscribers
Earliest in this view
Jul 8, 2026
The latest indexed work is over 30 days old. There may be a gap in what we hold.

Latest videos

  1. Video Β· Jul 25, 2026

    RuView: WiFi Can See Through Walls β€” Detects Your Heartbeat & Breathing with AI (opens the original)

    Excerpt Β· 3 min Β· 2,506 views by Sep 30, 2026

    Read excerpt

    RuView is the open-source spatial intelligence system that transforms commodity WiFi into a through-wall sensor. No cameras. No wearables. Just the radio waves already bouncing around your room. In this video, we break down how RuView captures Channel State Information from a $9 ESP32, processes it through a Transformer-GNN hybrid, and delivers contactless presence detection, breathing and heart rate monitoring, and full-body pose estimation β€” all running locally on a Raspberry Pi. What makes Ru

  2. Video Β· Jul 25, 2026

    Hallmark's Secret: What Makes AI Design Actually Good (opens the original)

    Excerpt Β· 6 min Β· 160 views by Sep 30, 2026

    Read excerpt

    Every AI-generated website looks identical. Purple gradient. Inter font. Centered layout. Your Claude Code and Cursor produce the exact same template as everyone else's. One developer found a way to break out of the training data. Six minutes that will change how your agent builds UI. ⏱ CHAPTERS 0:00 β€” The Problem: AI Slop 0:10 β€” Why Every AI Site Looks the Same 0:54 β€” Before vs After: Hallmark Comparison 1:16 β€” What Is Hallmark 1:24 β€” The Four Verbs: BUILD, AUDIT, REDESIGN, STUDY 2:00 β€” Twenty

  3. Video Β· Jul 13, 2026

    This Open Source AI Deploys a Team of Hackers to Find Your Bugs (opens the original)

    Excerpt Β· Neutral tone Β· 3 min Β· 230 views by Sep 30, 2026

    Read excerpt

    AI writes more of your code than you do. So who's actually checking it forvulnerabilities? Strix deploys a team of autonomous AI agents to hack your app, prove the bugs are real, and open pull requests to fix them. I cloned it, ran it, and dug into how the agent army actually hunts β€” and what it costs you. In this video: what Strix is, how its agents find and fix flaws, and my honest take after running it. Why does it beat a traditional scanner? And where does it fall short? GitHub: https://gith

  4. Video Β· Jul 13, 2026

    One Command, Full AI Pentest β€’ Strix CLI Walkthrough (opens the original)

    Excerpt Β· Neutral tone Β· 1 min Β· 150 views by Sep 30, 2026

    Read excerpt

    Want to run it yourself? One command. That's all it takes. ``` pip install strix strix run --target your-app.com ``` Boom. You've got a full AI pentesting team on your machine. πŸ”§ Python 3.10+ β€” works on Linux, macOS, Windows ⚑ CLI-first β€” `strix run` with live agent output scrolling πŸ”„ CI/CD ready β€” every pull request gets auto-scanned, fresh pentests on every code change πŸ“Š Structured reports β€” severity, location, working PoC, fix recommendations (not a wall of noise) Hook it into your pipeline a

  5. Video Β· Jul 12, 2026

    Strix β€” Not One AI. A Whole Hacker Team. (opens the original)

    Excerpt Β· Neutral tone Β· 1 min Β· 193 views by Sep 30, 2026

    Read excerpt

    Ever wish you had a team of hackers on demand? Strix gives you exactly that. Strix isn't one AI. It's a whole team of specialized agents working together: πŸ•΅οΈ Recon Agent β€” scans and maps your attack surface πŸ’₯ Exploit Agent β€” attacks with real hacker tools (browser, proxy, shell, sandbox) βœ… Validator Agent β€” proves every finding with a working exploit Each agent carries real hacker tools β€” a browser for web attacks, a proxy via Caido, a shell, and a sandbox. They share discoveries, chain vulnerab

Publishing over time

Last 90 days. Choose a month to open its work.

Recurring subjects

Named in the text we hold. One piece can cover several.

Audience

~272 subscribers

Measured Sep 19, 2026

Source's subscribers, not the number who saw an individual piece.

How this was measured

About this data

Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.

Identity or attribution wrong? Suggest a correction.

See coverage about Open Source Spotlight