7 Minute Security

7MS #726: Baby's First Hermes

7 Minute Security

Show Notes

Hello friends! I've been on a bit of an AI agent journey lately, and today I'm sharing my experience ditching OpenClaw and going all-in on Hermes — a self-hosted AI agent built by Nous Research. A Network Chuck video sold me on it, I wiped my Mac Mini (again), and baby's first Hermes adventure began! Here's what we get into today: Why I left OpenClaw — After getting the Mac Mini set up, OpenClaw left me feeling pretty meh: burning through API requests, random mid-conversation shutdowns, and a marketplace where the top listings were flagged as "potentially malicious." Hard pass. Network Chuck's five reasons Hermes rocks — His video summarized why Hermes stands out: (1) Nous Research has serious open source model cred predating OpenClaw, (2) more flexible persistent memory via markdown files + optional Honcho integration for building a profile of you over time, (3) a mission around humanistic and democratic AI, (4) a self-improvement loop where it writes its own skills after figuring things out, and (5) it just doesn't break — it feels like a product, not a project. The install — I used Claude to build a Mac Mini install guide from the Network Chuck transcript, and had Hermes up and running in about 15 minutes (one small Ollama hiccup aside). The install wizard lets you choose cloud models like Claude or ChatGPT, or go fully local with something like Gemma — I'm planning a hybrid setup with two Telegram bots. First real-world use: sitting in a truck running errands — With Hermes running on the Mac Mini and connected via Telegram, I asked it what it could do. It suggested Uptime Kuma for LAN monitoring — weirdly well-timed since I'd just been thinking about flaky IoT devices. I said "go install it," and it did — narrating its own troubleshooting out loud the whole time like a little robot intern. Remote access and Home Assistant — Had it install Home Assistant for smarthome control too, with plans to wire up TwinGate for remote access (it had a TailScale skill ready to fire in about two seconds, but I'm trying to keep VPN services consolidated). Daily digest via email — Hooked Hermes into a dedicated Gmail account and set up a 6 a.m. cron job that sends me a personalized morning digest: weather for my watched locations, recent breach/CVE news from select sites, and a summary of my favorite pentesting-focused Mastodon accounts. Needs tuning, but the first digest landed this morning and it's really good! The privacy angle — The real long-term win I see here is a hybrid model: feed raw, unsanitized pentest data to a local private model, let it analyze and sanitize, then hand off the clean version to a cloud model for deeper insight. Best of both worlds without the data exposure anxiety. Check out the Network Chuck video that started it all, and as always, if you're doing cool AI + security stuff, I'd love to hear about it. Find our pentesting services and training at 7MinSec.com, pentesting tips and scripts at 7MinSec.wiki, and if you want to support the show, head over to 7MinSec.club.

More from 7 Minute Security

Oct 17, 2026

Hey friends! Today I'm mashing two of our mini-series — eating the security dog food and my new obsession with n8n — into one big automation sandwich. I'll walk you through how Claude and a bunch of JSON files helped me build a self-updating asset inventory in Notion, firewall rule reviews, patch-ticket automation, and a push-button external pentesting and OSINT sweep. Basically, imagine I've cornered you at a party and won't stop talking about n8n.

Oct 9, 2026

Hey friends, it has been a long time since we talked about eating our own security dog food. However, I couldn't be more excited about some big ways we've moved the 7MinSec security program forward, including the release of our shiny new trust center. Enjoy!

Oct 2, 2026

Hey friends! I've got some personal and work travel this week, so I'm skipping the tangents (mostly) and jumping straight into something I'm nerdily excited about: Baby's First n8n. If you're not familiar, n8n is a platform for chaining together and automating just about anything your heart desires – and naturally, my heart desires automating away the boring parts of pentesting. In this episode I cover how I finally got hip to n8n (late to the party, as always), the Black Hills webcast that nudged me into it, how I fed Claude my Frankenstein "Swiss Army" pentest script and asked it to turn push-lots-of-buttons into push-one-button, the clever workaround for firing nmap, EyeWitness, and other tools like it (plus the safety gate I built so my automation can't go rogue and get me in trouble), a full OSINT recon fork powered by Project Discovery Neo that lands report-ready in SysReptor, and the many minutes of my life I get back from each OSINT run – and why that's the whole point. By the way, n8n is not a sponsor or partner – just something I like in case you like it too. Got cool n8n workflows of your own? I'd love to hear 'em!

Sep 25, 2026

Hey friends! Today's tale of pentest pwnage is brought to you by four hours of sleep, a large mint hot cocoa, and unhealthy levels of giggity. There's a very good reason for all three, but you'll have to hang out to the end of the episode to hear it. First, let's talk about an escalation path I've never walked before: A really tight network is a beautiful and terrible thing — this client went from a C-minus-ish environment in past years to one where I could find basically nothing. No local admin creds laying around, no poisonable protocols, no juicy file shares. Straight into "well, try harder" mode I went. Two new-to-me tools join the arsenal — one that gives you a second opinion on a big pile of pilfered files (Titus), and a one-shot PowerShell command (GPOZaurr) that hands you a report card for everything that's wrong with your GPOs (plus how to fix it). The finding I almost walked right past — an ADCS issue where only a tiny handful of objects could enroll. I said "all right, I'll move on." Then I came back to it, and saw I had missed the (kind of) obvious attack path. A certificate I couldn't authenticate with — and the 0xdf write-up that rescued me from my own gap in knowledge. Machine account quota was set to zero — so of course I tried to add a computer object anyway. Spoiler: it worked like a dirty shirt. The scheduled task trick rides again — my favorite "this is not going to get fixed" Windows behavior. I had a client push back on this one hard: "You had to know the password of the DA you abused, right? You stole the TGT, right?" Nope and nope. How to actually defend against all of this — some ADCS tightening, a checkbox on your high-privilege accounts that a lot of folks don't have ticked, and my new mantra: RDP with purpose. And then it's personal time — my oldest son Cam passed his national paramedic exam, and I am busting my buttons. Stick around for the whole story, including the 4 a.m. potty break that changed everything, the two health care professionals I'd like to hand a party invitation to in person (and why my wife says "I don't know about that, dude").

Sep 18, 2026

Hey friends! We've been on a bit of a Tales of Pentest Pwnage bender lately, so let's keep it rolling with Part 90. (And Mom, relax — this is not one pentest story chopped into 90 parts.) Today is less of an A-to-Z story and more a pile of tips and tricks pulled from a recent string of SCCM-flavored internals — plus a tangent about a video game and a little robot Claude and I built to get my life back. Multi-tier SCCM is having a moment — I've never administered SCCM a day in my life, but 2026 keeps dropping me into these split-role environments. Here's where I go to figure out my attack surface when all I've got is a low-priv cred. SMB signing on? Cool, I'll go around it — relaying from one SCCM box to the one with SQL on it, and the easy-button tool that vacuums the good stuff out of the database once you're there. (NAA creds, clear-text local admin passwords, install scripts with creds baked in…yes please.) Then relay the other direction — when the loot came back stale, a nudge from a Slack channel had me pointing the relay backwards, and I giggled like a little schoolboy at what popped out. Adding yourself to the local admin group: still weirdly undetected — an old episode of ours reminded me of an Impacket tool I don't see written up on many pentest blogs, and it slipped right by. My favorite cheat code hit a snag — the evil-scheduled-task-under-a-logged-in-DA trick kept coughing up permission errors, so I had to get creative. Plus the defensive recs I'm still trying to sharpen up — if you've got a better way to close this loophole, I'm all ears! Tangent: Halloween (the video game) and the lobby watcher — a million players and I'm still staring into the lobby abyss for 10 minutes at a time. So Claude and I built something that watches the screen and texts me when it's time to sprint back to the keyboard. It's not cheating. It's not! (The Texas Chainsaw crowd disagreed, loudly.)

See all episodes →