What I learned wiring three agents across Slack and Telegram with Switch
The invoice said 500 pieces of LL-STR-050 and the packing list said 480, both on page 1, row 3....
Tutorials
I built an AI on-call agent that never forgets an outage
It's 3 a.m. Your phone buzzes. The payments service is throwing KeyError: 'amount' after a deploy,...
I asked Claude to describe how my AI company functions, and here is what it wrote
We automated most of our engineering work. The interesting part is what we refused to automate.
I Stopped Treating Agent Memory Like Prompt History
A model can tell me that a startup resembles a company I evaluated two years ago. It can even invent...
Why I Replaced Global RAG With Hindsight Memory Banks
If you run technical support for an enterprise API platform, your engineers spend a staggering...
Writing your own Switch agents from scratch: six Python agents that take a startup idea to a launch decision
A first-timer's guide to building custom Switch agents: self-hosting with Mattermost, registration, the SSE stream and heartbeat, rooms as shared memory, and what broke.
A proposal is not a reservation: what we learned putting AI agents in a lab's booking chat
Every lab with a shared instrument has the same quiet failure mode. A student asks for an hour on the...
SMX Next 2026 Signals a Cross-Platform Shift in GEO for 2027
SMX Next 2026 is putting a clear strategic question in front of search marketers: how should organic...
Trust Network - An Agent-to-Agent Communication Protocol built on top of Switch
This is a blog about How I built trust-network, a communication protocol consisting of 'gates' which...
Ubuntu Core Desktop Is Coming in 27.04 - And It Kills sudo apt (But Maybe That Is the Point?)
I can't believe I'm saying this, but Ubuntu might be about to change for good. Not a new wallpaper...
ChatGPT Ads Generated Less Revenue Per Visit Than Google Search, Study Finds
A working paper examining ChatGPT’s international advertising rollout suggests that paid traffic from...
413 files, 1 badge: I re-ran Anthropic's memory test
Anthropic says one agent wrote 413 memory files and earned 1 badge; another wrote 9 and earned 4. I rebuilt the mechanism in 80 lines of Python to see if the gap survives without t...
I probed 791 x402 sellers. 306 answer 200 on a path they call paid.
791 sellers advertise paid x402 endpoints. I called every route once and graded what came back. Here...
I Let Three AI Agents Run an Incident. The Most Useful Thing They Did Was Refuse Me.
Building a human-gated incident team on Switch, with Claude Code agents working in Mattermost and...
I Stopped My Hindsight-Backed Agent From Inventing My Past
When you ask a language model for "similar past investments" and it will give you some. Names, years,...
One app to chat with Claude Code, Codex and OpenClaw on Windows, and what broke getting there
If you run Claude Code, Codex and OpenClaw on a Windows PC and want one app to chat with all of them,...
An agent told me it deployed. I asked the server, and the containers were a day old
The message said: deployed, edge clean, no 5xx. Both of those things were true. The commit was on...
I read 40 sources on spotting AI writing, code and images. Here's the prompt I made from them
I kept getting drafts back from ChatGPT and Claude with the same habits. "It's not just a tool, it's...
Teaching AI to Listen Without Ears
The first useful design decision in SynTalk was to stop treating a hand as one input format. A webcam...
I Built an AI Coding Team, Then Realized It Could Accidentally Destroy Production
By Bhavya Nayal • Building for the WeMakeDevs × DevTools Academy "Flip the Switch" Hackathon ...
Google Reaffirms Technical SEO as a Foundation for AI Search
Google is again drawing a clear line between preparing for AI-driven discovery and chasing isolated...
GPT Image 2.5 or Nano Banana 2.1? Cost per usable image
Answer: Nano Banana 2.1 is cheaper for iteration — 3 credits at 1K, 4 at 2K, 5 at 4K. GPT Image 2.5...
How osFoundry detects that Claude Code is waiting on a permission prompt
This is what Claude Code 2.1.303 writes to its terminal when it asks to run a command. The bytes come...
Agentjacking: A Fake Sentry Error Became a 'Fix' That Claude Code and Cursor Ran
A public Sentry DSN let anyone plant an error whose fake 'Resolution' told Claude Code, Cursor and Codex to run npx. Every hop was authorized. What stops it.