vivekyyy

vivekyyy on Hacker News

Re:Ask HN: How do you guys stop coding agents from acting out of line? on Tuesday September 01, 2026 @05:14PM (#49528299)

This is possible, but I just kinda feel like at that point I should just generate the migrations in CI. The point of having them out of CI (for me at least) was so we can manually write in upgrade/downgrade logic when the autogenerated migration is a little off

Re:Ask HN: How do you guys stop coding agents from acting out of line? on Tuesday September 01, 2026 @05:13PM (#49528278)

We use alembic for it, kinda curious what a hook config would look like here. Alembic files are pretty nice because they basically specify what sections of the revision should and shouldn't be manually edited, but agents routinely just ignore that lol

Re:Ask HN: How do you guys stop coding agents from acting out of line? on Thursday August 27, 2026 @05:22PM (#49471447)

I can add some basic configs, but it seems kinda provider dependent, so I'm gonna need to do this differently for Claude, Cursor, and Codex. It will also require a custom pre-tool use plugin to allow something like "AI can generate DB migrations via CLI but cannot handwrite them", which is pretty annoying when I need a separate one per model

Ask HN: How do you guys stop coding agents from acting out of line?

Posted by vivekyyy on Thursday August 27, 2026 @11:30AM
from the ask-hn dept.

I’ve been pretty annoyed recently with my coding agents doing stupid things that are anti-protocol. I can encode rules like “agents shouldn’t write to the README” pretty easily, but things like “database migrations must be autogenerated and not handwritten” are much harder to enforce. Anyone have any tools they use for this or tips for how to deal with issues like that?

( Read More... | 8 comments | 2 points | on HN )

Show HN: Rudder – Red-Green TDD Workflow for Verifiably Comprehensive Specs

Posted by vivekyyy on Wednesday August 26, 2026 @01:00PM
from the show-hn dept.

https://github.com/RudderCode/Rudder

Rudder is a local Codex/Claude Code plugin that captures your specs and your coding agent session history to force your agent to rewrite unit tests exclusively from your expressed intent. This turns unit test coverage into a proxy for what percent of the code actually comes from decisions you've made and lets you see how much of your agent's generated code is uncovered by your spec. After...

( Read More... | no comments | on HN )

Re:Cursor launches Origin, GitHub alternative on Tuesday August 18, 2026 @08:11PM (#49354782)

There's so many GitHub alternatives at this point, I just saw Dorsey is making one too. Why is that the trendy thing to build now?

Re:Ask HN: What Happened to Spec-Driven Development? on Wednesday August 05, 2026 @10:50AM (#49183801)

Hmm, yeah I guess the issue really is that the spec is not comprehensive enough to guarantee one specific implementation. If the spec really did translate one-to-one then the spec would really become a higher level language

Ask HN: What Happened to Spec-Driven Development?

Posted by vivekyyy on Wednesday August 05, 2026 @09:08AM
from the ask-hn dept.

For a minute, it seemed like SDD tools like Kiro, OpenSpec/SpecKit, Tessl were gonna be the future of coding agents. It seemed like a good idea to me at least, and it resolved a lot of the issues people were having with coding agents kinda just going off and doing their own thing.Nowadays though, it seems like all the SDD tools have kinda fallen off but all of the complaints remain—coding...

( Read More... | 4 comments | 3 points | on HN )

Re:Ask HN: What was your big failure? How did you get around it? on Tuesday August 04, 2026 @11:45PM (#49178376)

I’m guessing you may need to hear this: hold the line!

Sincerely, a guy who’s one year in.

Re:Ask HN: What is a good format for a tool to report data to a LLM? on Tuesday August 04, 2026 @11:38PM (#49178332)

In my experience, LLMs are really good with SQL. If you can structure your data into easily queryable SQL data, your LLM will be able to generate queries to get what it needs really effectively

Re:How to Pivot? on Tuesday August 04, 2026 @11:35PM (#49178323)

To be honest, I think the answer is in your question: find a real problem to solve. I know that sounds like a simplistic answer, but I think the best advice is usually the obvious answer

Show HN: I Repurposed Unit Tests to Show How Much Coding Agents "Improvise"

Posted by vivekyyy on Tuesday August 04, 2026 @12:33PM
from the show-hn dept.

https://github.com/RudderCode/Rudder

As someone who is using agents to code for me a lot lately, I was starting to feel more and more like I'm losing touch with the code I write--it's starting to feel like Claude's code rather than mine. I'm pretty sure I'm not alone in feeling this way, so I made Rudder as a first earnest shot at trying to fix this problem.Nowadays with AI coding agents, unit tests are pretty much dead code--your...

( Read More... | no comments | 2 points | on HN )

Show HN: Rudder – Own Your Agent's Code Again with Test-Driven Development

Posted by vivekyyy on Monday August 03, 2026 @12:07PM
from the show-hn dept.

https://github.com/RudderCode/Rudder

( Read More... | no comments | on HN )

Re:Ask HN: What's the Point of Your Startup? on Friday July 31, 2026 @03:21PM (#49127465)

Honestly speaking, I think it was always possible for someone to replicate what you do in your SaaS, but maybe just a little slower. IMO the value was always in the fact that someone else is willing to obsess about the problem so you don't have to.

For example, I could probably always have make a Quizlet clone for myself, but even with agents around I'd rather just use Quizlet than spend the time making my own flashcard app, even if I have to pay like $5-10 a month. Just food for thought