HoneyBadger-Designs

06/26/2026

I built an AI social media team that runs on your brand, not guesswork.

Drop in a your business markdown files — brand voice, services, offers, audience, compliance rules — and the system reads them as the source of truth. From there it researches current angles, builds the campaign, writes platform-specific posts and reel scripts, runs a compliance check, and hands off visual direction and images ready for Canva.

Every piece is built around one question: will this create a lead? Nothing goes live without a human approving it.

Brand in. Strategy, posts, reels, and DM funnels out.

We're booking a few weekly mowing spots right now in Church Hill, Kingsport, Surgoinsville, & Rogersville. Same crew, same day, every week.

Call or text us for a free, no-pressure estimate and we'll lock in your weekly spot.

We show up and do the job right every time.

423-430-7051

06/08/2026

I don’t trust AI email tools that rely on prompts alone.

“Be careful.”
“Ask first.”
“Don’t delete anything important.”

That’s not safety. That’s wishful thinking.

So I built Inbox Sentinel differently.

The safety is built into the code architecture — not just the prompt.

Before any risky inbox action happens, the system is designed to preview what will change, require human confirmation, and log the action afterward.

No silent bulk deletes.
No runaway AI cleanup.
No handing your Gmail account to an agent and hoping it behaves.

AI should help clean up your inbox.

It should not have unchecked power to destroy it.

Inbox Sentinel — AI email cleanup with guardrails built into the architecture.

06/08/2026

Most AI email tools rely too much on prompts.

“Be careful.”
“Don’t delete important emails.”
“Ask first.”

That is not enough.

I built Inbox Sentinel with safety in the actual code architecture — not just in the prompt.

Before any risky inbox action happens, the system is designed to show a preview, require human confirmation, and log what changed.

No blind bulk deletes.
No runaway AI cleanup.
No handing your Gmail account to an agent and hoping the prompt behaves.

This is the difference between an AI tool that sounds safe…

and one that is built to be safe.

Inbox Sentinel — AI email cleanup with guardrails built into the architecture.

05/25/2026

I don’t think AI inbox tools should require blind trust.

Inbox Sentinel previews every bulk action, requires typed confirmation for risky changes, and logs every write with before/after state.

No permanent delete scope.
Worst-case cleanup goes to Trash.

Triage with confidence.

https://sentinel.honeybadger-designs.com/

05/25/2026

Every AI inbox tool I tried scared me.
Not "the UX is bad" scared. Actually scared.😱
Granting an AI agent the full Gmail scope means it can delete anything, permanently. Letting it auto-reply on my behalf means one hallucinated sentence could go to my biggest client. "Smart cleanup" means trusting a model I can't audit with the most sensitive system I own.
So I either lived with 14,000 unread emails — or handed over the keys.
That gap is why I built Inbox Sentinel.
Three rules baked into the architecture, not the prompt:
→ Preview every bulk action before it fires. You see the count. You see a sample. → Confirm with a typed string. "yes" for small. "yes trash 43" for big batches. → Log every write to an immutable Postgres table with before/after state.
The full-mailbox Gmail scope that allows permanent delete? Structurally banned! Worst case for any cleanup is Trash, where Gmail keeps it for 30 days.
It's still early. I'm the only one building it. But I'd rather ship a tool you'd actually trust your inbox to than another autonomous agent that sounds impressive in a demo.

https://sentinel.honeybadger-designs.com/

What's the AI feature you almost trusted, then backed away from? Comment below ⇓

02/18/2026

⚠️ Window Users Update ASAP!!

🚨 Microsoft just dropped 58 patches. 6 zero-days. All 6 actively exploited.

February 2026 Patch Tuesday is HERE and it is intense.

Let me give you the 60 second breakdown.

The scary part?

Three of these zero-days can chain together.

Windows Shell bypass ➡️ Word bypass ➡️ SYSTEM privileges.

That is a full attack path. From clicking a link to total compromise.

🔴 CVE-2026-21510: SmartScreen? Bypassed. Click a bad link, no warning, no prompt. Just ex*****on.

🔴 CVE-2026-21513: MSHTML bypass. Open a malicious HTML file and your security controls go silent.

🔴 CVE-2026-21519: Desktop Window Manager escalation straight to SYSTEM. Full control.

And here is a particularly gross one:

📧 CVE-2026-21511: Outlook spoofing via the PREVIEW PANE.

Per MSRC: Your users do not even have to click.

Just viewing the email in Outlook can trigger it.

This sophisticated... you know the threat actor behind this is not messing around.

Here is your action plan:

✅ Patch NOW. Not this weekend. Today.
✅ Check Outlook security. Preview pane = attack vector.
✅ Watch for IOCs on these CVEs.

This is the kind of Patch Tuesday that separates the "we will get to it" teams from the "we are on it" teams.

Which one are you?

All the details from Microsoft is here: https://msrc.microsoft.com/update-guide/releaseNote/2026-Feb

02/10/2026

SOOO.... I have had the Badger Bot live for a couple months and the response has been fabulous!! Preping to build Badger Bot 2.0, 2.0 will be for higher theory, critical thinking, and a database to hold and remember what each client has asked, and build its own knowledge base from the client. It will also use various AI models. Working towards building the famous J.A.R.V.I.S. (Just A Rather Very Intelligent System) for Badger Bot 3.0. Smash that Follow button to watch how my journey grows and develops.

https://honeybadger-designs.com/

.

12/20/2025

I stopped by to check in with Santa. He is busy watching the weather and checking his list.

12/10/2025

🚀 New Chatbot Alert – I Need Your Help!

I just finished coding a brand-new chatbot, and I’m looking for people to test it (a.k.a. fuzz it, break it, confuse it 😄) and tell me what you think.

👉 Try it here: https://honeybadger-designs.com/

I’m especially looking for feedback in English and Spanish:

How well does it understand your questions?

Where does it get confused?

What answers feel helpful vs. off?

What would you improve?

The chatbot is still a work in progress and doesn’t have all my business info loaded yet, but the core is solid. The cool part?
For my clients, the system prompt and knowledge library are fully controlled through a CMS customer dashboard in the websites I build — so they can tune what their chatbot knows and how it behaves without touching code.

If you have a few minutes, please jump in, ask it anything about HoneyBadger-Designs, and drop your feedback in the comments or DM. Your help means a lot! 🙏💬

https://honeybadger-designs.com/

11/15/2025

Your website shouldn’t just “look nice.”
It should light a fire under your leads. 🔥

If your site isn’t:
– Showing up on Google
– Loading fast
– Making your offer stupid-clear
– Giving people ONE obvious next step…

…it’s not a website. It’s a pretty business card.

Ready to turn your site into a 24/7 salesperson instead of a silent brochure?
Comment “FIRE” and I’ll send you details.