breckenedge

It adds some abstractions and a plugin architecture for extensions. I checked it out but ended up not using it. I run my own harness of harnesses over tmux centered around tasks rather than sessions.

Right. Can’t wait to see what all their competitors do to catch up.

Lithium fuel can melt asphalt seams

Right. Y’all are talking about two different types of multitasking.

I mean, you can also use agents to listen to users and handle issues automatically.

But I guess you mean the discovery part? Yea, agents are not going to totally replace that anytime soon. Though, Claude voice mode is pretty good these days, so I’m sure it’s on their mind.

Not sure what you mean by “stay the course.” I think the winners will be those with the most effective marketing strategies. Surely that will require automation.

Stepping on each other how? I find each agent needs to spin up N worktrees for each task, and I use hooks to prevent them working on the same substrait.

You’re not running tests in CI? The agent should only run the minimum tests necessary and then idle and wait for feedback from a CI run.

Different sort of multitasking.

Maybe I didn’t give it enough of a shot, but I fond herdr was solving the wrong problem for me. I want to manage shipping work, gtd, not a bunch of agent sessions. The sessions are an implementation detail that I prefer jumping into only when something is wrong.

How many simultaneous cursor windows are you running? Have you hit RAM limits at some point?

Understand when you’re in a meritocracy versus a plutocracy. Outshining the favorite son can be harmful.

Before we were married, my future wife and I played The Witness together on my iPad. Fond memories, hope this gets a mobile release too.

Yea. I haven’t opened VS code in I don’t know how long. Among other things, I just shipped a full mobile app for Android and iOS and never had to touch XCode or Android Studio.

The factory must grow.

Just about 90 PRs in that dataset to review.

This is all code review runs via OpenRouter with a Pi harness, and it’s totally possible there are shenanigans going on elsewhere.

Yesterday, I ran an identical bug identification dataset from two weeks ago, saw a 50% drop from a few weeks ago, putting Sol on the same level as Luna. Sol had been finding 40-50 bugs per set, then dropped to 25, matching Luna’s performance. Not enough to establish a pattern, but enough to raise eyebrows.

Our review workflow is public if you want to peruse it, dataset isn’t. The process isn’t really stabilized yet either as I have to balance running this against limited budgets.

https://github.com/BiggerPockets/.github/blob/main/.github/w...

Do these evaluations get re run a few weeks after launch? I started doing that yesterday for our internal dataset and found Sol’s performance had regressed to be equal to Luna’s. Granted this was one run, but something I’m becoming more concerned about, the model providers want to quickly prove they’re the best, people switch to them, then they pull the rug.

Yea I definitely keep engineering skills and org skills separate.

My approach is different. Skills I write mostly use Python to save on the agent needing to run its own loops.

I use these loops to monitor the CI build and PR approvals rather than having the agent poll, and even Opus gets the commands wrong enough to make it worth it.

Last week I wired up a skill for the agent to share screenshots in PRs via specific S3 buckets and AWS CLIs. Again, the agents guess at the right commands often enough to make it worth being explicit.

Sure these could have gone in CLAUDE.md, but not every agent needs the context.

And at the company level, I can push skills to everyone’s Claude via the Teams function, they don’t need to edit configs or even know what a skill is.

Those terms became way too ambiguous to actually mean anything. Every team does it different.

Yes, and not just code. It’ll be Gnosticism all over again.

HN reply bot

I get it doesn’t appeal to you. Why do you think it shouldn’t appeal to anyone?

Service area is still miniscule.

Yes, I fetch Jira tickets and epics via API call in the reviewer’s action. Both Codex and Claude can then reference it. Source is public here:

https://github.com/BiggerPockets/.github/blob/main/.github/w...

Still a bit nicer to just do these yourself rather than reusing someone else’s project.

Like a junior engineer, I find the models to be too ambitious and unable to steer themselves at a high level yet. What I’ve done to address this is prompting the model to break down its plans into more atomic steps. For whatever reason, they’re still lazy at planning.

Programming is human. Obviously there will be cultures (and cults) that spring up and die down all the time. I probably pissed off a few colleagues over the years by rejecting their arguments because they were “appeals to authority” that I did not recognize as authorities in the particular problems we were solving at the time. Hate cargo cult programming, but admit I fell into that trap often enough in my early years. Gotta start somewhere.

Oracle overcommitted on behalf of OpenAI, taking on debt to meet their obligations. Not that healthy companies are not also doing ridiculous layoffs, but in this case, Oracle does have to do some drastic things to get out of this tailspin.

Fyi the homepage is too small for my mobile browser (chrome, iPhone 15 pro) and appears to not support scrolling

It’s amazing how often I’ve discussed this from the perspective of “binary search”: you gotta make big jumps before you do narrow jumps.

Same for being stuck in local maxima: getting out of that requires big changes, not small-step optimizations.

References Claude 4.7 … written June 8 … ok cool can stop reading.