Notes from Claude Code Curious #2 — including conversation with Conner from Claude Code. Hosted at LocalGlobe, London, February 11, 2026.
Last week, 100 close personal friends (Claude Code Curious attendees) and I had the opportunity to quiz Conner from the Claude Code team in a live Q&A.
Because he sits on the Product team, he was able to bring a really broad perspective, with some good advice for anyone hoping to level up their use of Agents.
Sign up here to join us next time on 11 March!
One big trick: agree to disagree
“If you can put two agents at each other with opposite goals — so think kind of like Adam and Eve, where one’s the coder and one’s the evaluator, and they actually have this contract with each other — one’s writing the code, it stops, one then reviews the code, it QAs it, it gives the coder feedback. They’re basically at this contract where the cycle can’t end until the evaluator is happy with the coder’s work.”
Love this. In the human work world, its true that teams are greater than the sum of their parts, and it follows this would be correct with agents.
Of course the nuance becomes in how you compose the team, how you manage it, where you draw the lines and what’s allowed. What is too much disagreement?
Conner expanded:
“If you have a small task and you have six people debating implementation details, that’s just a waste of time. You should have one single agent working on it. Just like you should have one developer working on it.”
“The team lead would actually just terminate child teammates or child agents when they were disagreeing too much or refusing to do a task because their opinion was so strong.”
Agents work while you sleep
Conner described an internal setup at Anthropic where Claude handles the full cycle from Slack to shipped code:
“Claude would search Slack, find issues, add them to Asana. A teammate on the agent team would then pull that task in, push a fix or add a feature, commit it to GitHub, initiate a pull request, get it merged, back to Asana, mark it as done. Every day you go in the morning, you can see what’s been done, what’s been shipped.”
Twenty-four-hour-plus tasks are now possible using sub-agents and stop hooks. We’re in a time where most examples from people like Conner still reference coding. But I can’t help but feel it may be more exciting to describe real work.
Ask yourself the question: when you come in in the morning, what work would you wish was just done and waiting for you to review?
Maybe your Claude agents are like a team in a different timezone – time to give them Australian accents!
Three simple tips: Abuse Skills, Multitask, Compact early
Conner gave two specific tips for getting better results — and neither was about prompting.
“Abuse skills. If you’re doing something more than once a day, make it a skill. Just take the time to do it. You can have Claude make it for you.”
Second, multi-task. The model is fast enough that your bottleneck is you, not it.
“It can be difficult at first but force yourself to work on five things at once. Basically, every developer on the Claude Code team is working on around, like, five things at once.”
Personally I’ve started to dedicate an entire Mac desktop to different Claude Code chats, all on –dangerously-skip-permissions, and working away on whatever comes into my head on a whim.
Then, one I’m particularly guilty on: don’t let your context rot.
“Compacting is bring heavily studied right now and we’re investing a lot of time in it. But compacting early is better than compacting late. If you compact at 20% versus 80% on the same task, you actually have better results at 20%. So I’d recommend compacting early.”
But his preferred approach is to avoid compaction altogether:
“This is why we’re big plan mode advocates. You basically get your entire prompt together, polished, it’s good to go, and then you clear all prior context and just hit go. I think that has the best results.”
Everything successful started with one person’s itch
Conner described Anthropic’s product development as almost entirely bottom-up.
“That is true for Claude Code. Was true for Cowork. That was true for our Excel plugin, the person who built that was an investment banker previously. It’s all bottoms up. I think for anything successful today that we currently offer, an IC had an idea.”
How is this relevant to Claude Code Curious attendees? I think you should follow the same model. The future is here, it’s just unevenly distributed. But nothing is really stopping you leaping in, and whatever might fit that category is more easily dismissed than ever by using the tools themselves.
–
Join us again on 11 March for another Q&A and a great crowd of curious Clauders.