Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GitHub Agentic Workflows

github.github.io

Post reply on HN
Print view View original

141–149 of 149 posts

Re: GitHub Agentic Workflows

#141

Post by onionisafruit »

catlifeonmars wrote: post #136
threecheese wrote: post #87

Earlier quoted context omitted.

For the first, I think maintaining package-add instructions is table stakes, we need to be opinionated here. Agents are typically good at following them, if not you can fall over to a Makefile that does everything. For the second, I totally agree. I continue to hope that agents will get better at refactoring, and I think using LSPs effectively would make this happen. Claude took dozens of minutes to perform a rename…

Does the agent have a way to interact with the lsp?

I don’t know about other lsps, but gopls has an -mcp flag that makes it run an mcp server. There’s also a jetbrains plugin for claude that gives claude the ability to use a subset of your jetbrains IDE’s features.

I usually have both of those configured when using claude on Go repos, and I still have the same frustrations as the comments above. Gopls has symbol search, but claude almost always uses grep to find uses instead.

Re: GitHub Agentic Workflows

#143

Post by nine_k »

raxxorraxor wrote: post #130
nine_k wrote: post #115

Earlier quoted context omitted.

How much does it cost you?

I do it similarly and it only costs me my working time. I do some of these things outside my official working times for free, but that is because I like the topic and like to have a good deployment pipeline. But I doubt it is in any way more significant time investment than administration of Github. In the end you need to write your deployments scripts yourself anyway, which takes the most time. Otherwise for install…

I mean, agents don't run for free for any significant time. Tokens cost money.

Re: GitHub Agentic Workflows

#144

Post by resquawk »

catlifeonmars wrote: post #140
andrewmcwatters wrote: post #32

There's a lot of hate in this thread, but there are plenty of engineers chomping at the bit for autonomous workflows, because browser-use isn't there yet, and cloud expenses from major providers are also unappealing with so much relatively powerful local compute.

It’d be fine if they included a big disclaimer at the top that this is beta software and they’re not liable for blah blah blah, but without such a disclaimer it’s reasonable to assume the software is ready for production. I think much of the hate is coming from GH misrepresenting its software and people being surprised by the many minor bugs.

Thanks! We've moved the disclaimer up the page :)

Re: GitHub Agentic Workflows

#145

Post by onionisafruit »

gloosx wrote: post #131
onionisafruit wrote: post #53

I noticed this unusual line in go.mod and got curious why it is using replace for this (typically you would `go get github.com/Masterminds/semver/v3@v3.4.0` instead). replace github.com/Masterminds/semver/v3 => github.com/Masterminds/semver/v3 v3.4.0 I found this very questionable PR[0]. It appears to have been triggered by dependabot creating an issue for a version upgrade -- which is probably unnecessary to begin w…

I like how it accumulated 3 such replacements before finally getting fixed as a reaction to this comment with PR 14543[0], but after review, two "fix unit tests" commits were added, one of which replaces claude with copilot and the second one messing up the docs markdown, getting merged after that. Agentic workflows are the battle... https://github.com/github/gh-aw/pull/14543

I’m glad they are dogfooding this in public because it is a good indicator to stay far far away

Re: GitHub Agentic Workflows

#146

Post by raxxorraxor »

nine_k wrote: post #143
raxxorraxor wrote: post #130

Earlier quoted context omitted.

I do it similarly and it only costs me my working time. I do some of these things outside my official working times for free, but that is because I like the topic and like to have a good deployment pipeline. But I doubt it is in any way more significant time investment than administration of Github. In the end you need to write your deployments scripts yourself anyway, which takes the most time. Otherwise for install…

I mean, agents don't run for free for any significant time. Tokens cost money.

I meant the CI agents, not AI agents. These are just runners that execute the stuff that needs to be done for deployment/testing or general CI. These rarely call AI agents because these tasks need to be deterministic. If so, you probably want to call a local model under your control running on your own compute power.

Re: GitHub Agentic Workflows

#147

Post by catlifeonmars »

onionisafruit wrote: post #141
catlifeonmars wrote: post #136

Earlier quoted context omitted.

Does the agent have a way to interact with the lsp?

I don’t know about other lsps, but gopls has an -mcp flag that makes it run an mcp server. There’s also a jetbrains plugin for claude that gives claude the ability to use a subset of your jetbrains IDE’s features. I usually have both of those configured when using claude on Go repos, and I still have the same frustrations as the comments above. Gopls has symbol search, but claude almost always uses grep to find uses…

Didn’t know about the go lsp builtin mcp server. That’s neat!

Does preventing the agent from using a shell help at all with the grep issue?

Re: GitHub Agentic Workflows

#148

Post by TZubiri »

throwup238 wrote: post #123
TZubiri wrote: post #121

Earlier quoted context omitted.

Correct. First of all, any subdomain system domain is already a bit phishy because you need to somehow parse whether github.io is officially part of github.com and not say something like git-hub.xyz by a phisher or whatever new TLD there. These things are used by sysadmin/project pairs that can't budget 1$/month for a domain name, so it's 100% a security/price tradeoff. Second of all, the actual domain host is publis…

I get your general objections, but not in this specific case. Github has been using Github.io for pages since 2013 and it's been the de facto developer platform at least as long (and all other developer tools follow the same pattern when publishing user generated content). Unless GH has a massive vulnerability that hasn't been discovered yet, no one is publishing to *.github.github.io except for the official Github o…

Would it really make a difference if they just added a CNAME from foobar.github.com to point at github.github.io?

Yes, that would help, but it's not very discoverable.

I think a certificate mechanism would be much more appropriate for that.

The SSL certificate should be emitted for github.com and github.io

Of course since github.io is rented out, it doesn't make sense. But if you ever have an alias, that's the way to do it, if I get a link to getproduct.com and it gets redirected to product.com I can check the cert and see that it was issued for both domains.

Re: GitHub Agentic Workflows

#149

Post by hmokiguess »

TZubiri wrote: post #127
hmokiguess wrote: post #15

Earlier quoted context omitted.

How is it not confirmed? GitHub cannot use their own product? Them using GitHub pages changes something? I don’t get it

I see a lot of people confused, and it is confusing. Here's my best take at clarifying the issue for you: It's as if Google sent you an official email from an @gmail address. Like "gmail-invoices@gmail.com" Surely it would look suspicious, and if it turns out it is official, it doesn't somehow mean there's no issue, if anything it's worse because it untrains users' security protocols. Personally I'd ignore anything t…

I don't think this is the same thing though, in their documentation the subdomain has to be for the owner or organization, so if it's `github.github.io` then it belongs to the `github` organization [1]

Though I guess I do get your point, meaning that this leads to people trusting the domain name if they are uneducated about how GitHub Pages works.

[1] https://docs.github.com/en/pages/getting-started-with-github...

Post reply on HN