I’m seeing discussions on the planning and development cycle with agents. I applaud them because those were the missing pieces in project planning when working with other humans. The way we work with agents is exactly how we should work together. We would accomplish so much.
Is silence golden in IT?
When a tool goes quiet, is everything fine, or is the tool broken?
This week I built an AI agent with Claude Code that never lets silence pass for "all clear." 🧵
hey @1Password you should have 2 KPIs:
- avg credentials per workspace
- avg time to fill a password
please just optimize for that
it's starting to get crazy how often i have to reauthenticate, unlock, or reopen the app because it's stuck
Demolished dinner & it’s time for dessert as the World Series plays in the background.
Made a GPT tailored to my network automation journey giving me exercises based on where I am in the books. Gives me hints as I request it. The GPT analyzes my code and gives me ways to improve
People call this vibe coding. It feels more like writing a runbook for someone smart who works very fast.
What would an AI agent have to prove before you let it near your network?
Still in progress: our first agent is hardening with restore drills and self-monitoring. Guardrails on the new one just started, including tests it has to refuse.
@jgwifi That's what I'm working towards! I'll be sharing lots more about it. One of the next agents I want to tackle using a new workflow will be the Wi-Fi agent. But that one will be harder to build because I need to take my knowledge from my head and put it into an LLM.
I'm at the age where staying home is what I want to do on the weekends. And instead of building tech, I'm building a place I want to stay at and be happy.