Building software and teams with agents and humans. Founder of Connsulting, co-host of @managerxbuilder
Washington, DC
Joined May 2013
- Tweets620
- Following32
- Followers43
- Likes75
Pinned Tweet
Rebooting my X. Expect a lot of AI talk. Planning to discuss how and what I'm building these days.
Brian Conn retweeted
Brian deleted all of his custom sub-agents this week. He still runs vanilla sub-agents constantly, because a test writer shouldn't know what the implementer did.
managerxbuilder.substack.com…
Brian Conn retweeted
Brian's personal wiki ranks sources by trust. His own words count most, then his calls, then the podcasts and newsletters he trusts. Random web results never get in.
Brian Conn retweeted
Tomorrow on Manager Builder, Kunal and Brian debate generalized vs specialized agents.
"Most of our best practices are just best practices for humans."
Brian Conn retweeted
In planning, Claude should ask what you want to do and then stop. Once it hands you a recommended option, it's too easy to just hit Enter.
managerxbuilder.substack.com…
I think TTS is included in the @grok unified supscription. It works.
Going to see what else works in their API with the sub...
Brian Conn retweeted
One-shot vs progressive on the same monitoring stream.
One-shot: four false pages, missed the leak, a couple of cents.
Progressive: one page, no false positives, about 50 cents.
managerxbuilder.substack.com…
My take on /grill-me:
Grill me on functional and architectural requirements. You spike technical hurdles to validate or refute my decisions and de-risk the planning process.
I wasn't using @cursor before The Grand Unification, but man, no included usage for other models sucks.
Used to use Cursor a lot. Good for the team, but it's officially been eaten up.
I moved off MCP using mcporter today.
I have too many projects with too many model providers with different configs. Tired of keeping them in sync. mcporter got me CLI-like interfaces.
Now I can use whatever model in any project without worrying about MCP compatibility.
Started building my own personal agent benchmark yesterday. Lots of tweaking today. Deep dive episode coming soon to @managerxbuilder.
Brian Conn retweeted
Asked a fast classifier to make one call per event. Four false alarms, and it missed the memory leak.
Brian Conn retweeted
AI slop can cost you a client.
Manager: keep a quality gate when models write the code.
managerxbuilder.substack.com…
Is all written content suspect now? Slack, email, tickets. How do you verify anyone behind the message understands what's written?
I'm leaning towards calls more these days for high-level design and architecture.
Some of these calls are the most fun I have all day.
I stopped using @grok 4.7 for my local /implement commands. Quality isn't there compared to Sol 6.1 or Opus 5.5. I still use Grok 4.7 as an adversarial reviewer.
Looking forward to Grok 4.8.
Brian Conn retweeted
Why personal software matters: remove friction in your own workflow, and get a low-risk lab to try AI before work.
Builder.
managerxbuilder.substack.com…
Brian Conn retweeted
Squarespace had no API, so I was the human in the loop adding zero value.
One day-long side project with AI later, the site runs on Astro 5 + Cloudflare Pages.
Builder take on personalized software.
Brian Conn retweeted
Are you leading Claude, or is it leading you?
Manager take: lead the bots instead of being led by them.
managerxbuilder.substack.com…
