You're not using your agent
You ARE its loop
It writes, you read. It says "done", you check. Then you type again
Prompt, context, harness, loop, graph: the full agent stack in one course, and how to stop babysitting ↓
If you're still AI coding by babysitting a bunch of terminals you should really look into workflows. It took me a few tries but my agents now work non-stop for days with occasional steering.
I now spend most of my time focusing on design and architecture while my agents code.
19
Brandon Kindred retweeted
We've given all Pro, Max, Ultra, and Team users one free full reset, across all models and limits. Use it anytime before December 10 at 2:00PM PT.
We recently fixed a few small issues that used more limits than intended, and the Fable resets didn't do much for those of you who'd already hit your weekly limits for all models. While these issues didn't affect everyone, we want to make up for the trouble.
You'll find it in Settings > Usage, or at clau.de/reset.
I tried the vibe coding thing and it did not go well. Nothing worked right
I've been devibing it using my usual method for unfucking legacy code
16,743 issues later and its finally sort of working and maintainable. I think it'll take another 16k issues
Don't vibe code folks
1
62
Unpopular opinion. Companies often hire devs that they think are engineers. If you haven't figured out which one you hired, AI is about to make it obvious.
Hint: if your folks are unable to show real value from AI coding tools or refusing to use AI...😬
20
Brandon Kindred retweeted
New Anthropic research: How AI agents shape experts' judgment calls
When experts delegate work to an agent, do they still make the key decisions? We tested this in a randomized trial with 50 professional data analysts.
My favorite Claude Code behavior is when it complains that there have been a lot of code review rounds and every time I find a ton of issues so we should just merge the PR?
It really is like a junior engineer...
5
2
75
Brandon Kindred retweeted
We're donating Agent Skills to the Agentic AI Foundation (@AgenticAIFdn), the same foundation we donated MCP to.
Since December 2025, 70+ agent platforms have integrated the standard, including ChatGPT, Cursor, GitHub Copilot, and VS Code.
Here's what that means for you:
I learned to write software on microcontroller with 512kb memory and an 80ms max clock cycle.
You slop it out on computers with literally millions of times more capacity and somehow still create slow software
We're not the same.
25
Brandon Kindred retweeted
Code Review is now generally available for Claude Team and Enterprise plans.
Reviews are now faster, catch more of the bugs that matter, and admins can pick from Quick, Standard, or Deep review levels for each repo.
Hot take
everyone saying that software engineering is solved and AI is going to destroy all the jobs are oblivious to the fact that a significant portion of the world still runs on mainframes from the 70s and almost none of it is documented.
Chill folks...we have time
19
Everyone talks about Agents but why does no one ever talk about Akka.io? I feel like if you build agentic systems akka should be talked about way more often than it is.
17
Today my agents went to work and burned through 3000 GitHub actions minutes in less than 5 hours. Yikes. Claude had me laughing when I asked for alts
"Pulling your actual CI run history to see where the minutes are going — 20,849 runs on record, so this is going to be fun."
1
3
63
Brandon Kindred retweeted
Claude Managed Agents dynamic workflows are now available in public beta. It's a new type of multiagent orchestration, built for your most ambitious workloads.
A lead agent writes a plan that runs across many agents in phases, combining the results at the end.
Everyone seems to think that it's impossible to know where tech is going now that LLMs exist. I don't get it. The industry has collectively gone back to desktop apps and CLI. You want to know where things will go, learn how the industry progressed since 2000.
17
Hot take: if everyone stopped treating LLMs like super apps the tech might actually live up to the hype.
13
Everyone saying that we have reached AGI are either not bright enough to understand what intelligence is or has a vested interest in you believing nonsense.
I'll believe that we are approaching AGI when LLMs can proactively identify context gaps and proactively gather more info
12
Everyone needs to stop with the posts about whether or not engineers are needed now that LLMs can write almost all code.
If you think engineers are obsolete then prove it. Show us your work or shut up.
15





