<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: dakshgupta</title><link>https://news.ycombinator.com/user?id=dakshgupta</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 22:35:39 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=dakshgupta" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[How to build a code review agent]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/how-to-build-a-code-review-agent">https://www.greptile.com/blog/how-to-build-a-code-review-agent</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=50091917">https://news.ycombinator.com/item?id=50091917</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 14 Oct 2026 17:52:56 +0000</pubDate><link>https://www.greptile.com/blog/how-to-build-a-code-review-agent</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=50091917</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50091917</guid></item><item><title><![CDATA[New comment by dakshgupta in "PR spam today looks like email spam in the early 2000s"]]></title><description><![CDATA[
<p>Apart from the job-related stuff others have already said, there is a bit of novelty/bragging rights in landing a PR into a major open source project.</p>
]]></description><pubDate>Wed, 24 Jun 2026 16:58:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=48662700</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=48662700</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48662700</guid></item><item><title><![CDATA[PR spam today looks like email spam in the early 2000s]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/prs-on-openclaw">https://www.greptile.com/blog/prs-on-openclaw</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48660579">https://news.ycombinator.com/item?id=48660579</a></p>
<p>Points: 265</p>
<p># Comments: 155</p>
]]></description><pubDate>Wed, 24 Jun 2026 14:32:12 +0000</pubDate><link>https://www.greptile.com/blog/prs-on-openclaw</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=48660579</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48660579</guid></item><item><title><![CDATA[New comment by dakshgupta in "TREX: An AI code reviewer that runs your code"]]></title><description><![CDATA[
<p>Not yet - but we want to do this. Similarly true for the ephemeral unit tests that greptile writes.</p>
]]></description><pubDate>Wed, 17 Jun 2026 20:40:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=48576552</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=48576552</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48576552</guid></item><item><title><![CDATA[TREX: An AI code reviewer that runs your code]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/trex-code-execution">https://www.greptile.com/blog/trex-code-execution</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48571851">https://news.ycombinator.com/item?id=48571851</a></p>
<p>Points: 60</p>
<p># Comments: 11</p>
]]></description><pubDate>Wed, 17 Jun 2026 15:29:01 +0000</pubDate><link>https://www.greptile.com/blog/trex-code-execution</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=48571851</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48571851</guid></item><item><title><![CDATA[Slop is not necessarily the future]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/ai-slopware-future">https://www.greptile.com/blog/ai-slopware-future</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47587953">https://news.ycombinator.com/item?id=47587953</a></p>
<p>Points: 305</p>
<p># Comments: 484</p>
]]></description><pubDate>Tue, 31 Mar 2026 14:32:26 +0000</pubDate><link>https://www.greptile.com/blog/ai-slopware-future</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=47587953</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47587953</guid></item><item><title><![CDATA[New comment by dakshgupta in "There is an AI code review bubble"]]></title><description><![CDATA[
<p>The signal-to-noise ratio problem is unexpectedly difficult.<p>We wrote about our approach to it some time ago here - <a href="https://www.greptile.com/blog/make-llms-shut-up">https://www.greptile.com/blog/make-llms-shut-up</a><p>Much has changed on our approach since then, so we'll probably write a a new blog post.<p>The tl;dr of what makes it hard is 
- different people have different ideas of what a nitpick is
- it's not a spectrum, the differences are qualitative
- LLMs are reluctant to risk downplaying the severity of an issue and therefore are unable to usefully filter out nits. 
- theory: they are paid by the token and so they say more stuff</p>
]]></description><pubDate>Tue, 27 Jan 2026 07:00:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=46776408</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46776408</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46776408</guid></item><item><title><![CDATA[New comment by dakshgupta in "There is an AI code review bubble"]]></title><description><![CDATA[
<p>Thanks! We go over that on many other pages. Here are some:<p><a href="https://www.greptile.com/benchmarks">https://www.greptile.com/benchmarks</a>
<a href="https://www.greptile.com/greptile-vs-coderabbit">https://www.greptile.com/greptile-vs-coderabbit</a>
<a href="https://www.greptile.com/greptile-vs-bugbot">https://www.greptile.com/greptile-vs-bugbot</a></p>
]]></description><pubDate>Mon, 26 Jan 2026 18:54:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=46769906</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46769906</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46769906</guid></item><item><title><![CDATA[New comment by dakshgupta in "There is an AI code review bubble"]]></title><description><![CDATA[
<p>I agree that none perform _super_ well.<p>I would argue they go far beyond linters now, which was perhaps not true even nine months ago.<p>To the degree you consider this to be evidence, in the last 7 days, the authors of a PR has replied to a Greptile comment with "great catch", "good catch", etc. 9,078 times.</p>
]]></description><pubDate>Mon, 26 Jan 2026 18:51:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=46769868</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46769868</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46769868</guid></item><item><title><![CDATA[New comment by dakshgupta in "There is an AI code review bubble"]]></title><description><![CDATA[
<p>2. There is plenty of evidence for this elsewhere on the site, and we do encourage people to try it because like with a lot of AI tools, YMMV.<p>You're totally right that PR reviews go a lot farther than catching issues and enforcing standard. Knowledge sharing is a very important part of it. However, there are processes you can create to enable better knowledge sharing and let AI handle the issue-catching (maybe not fully yet, but in time). Blocking code from merging because knowledge isn't shared yet seems unnecessary.</p>
]]></description><pubDate>Mon, 26 Jan 2026 18:42:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=46769760</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46769760</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46769760</guid></item><item><title><![CDATA[New comment by dakshgupta in "There is an AI code review bubble"]]></title><description><![CDATA[
<p>> Independence<p>It is, but when a model/harness/tools/system prompts are the same/similar in the generator and reviewer fail in similar ways. Question: Would you trust a Cursor review of Claude-written code more, less, or the same as a Cursor review of Cursor-written code?<p>> Autonomy<p>Plenty of tools have invested heavily in AI-assisted review - creating great UIs to help human reviewers understand and check diffs. Our view is that code validation will be completely autonomous in the medium term, and so our system is designed to make all human intervention optional. This is possibly a unpopular opinion, and we respect the camp that might say people will always review AI-generated code. It's just not the future we want for this profession, nor the one we predict.<p>> Loops<p>You can invest in UX and tooling that makes this easier or harder. Our first step towards making this easier is a native Claude Code plugin in the `/plugins` command that let's Claude code do a plan, write, commit, get review comments, plan, write loop.</p>
]]></description><pubDate>Mon, 26 Jan 2026 18:39:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=46769696</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46769696</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46769696</guid></item><item><title><![CDATA[There is an AI code review bubble]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/ai-code-review-bubble">https://www.greptile.com/blog/ai-code-review-bubble</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46766961">https://news.ycombinator.com/item?id=46766961</a></p>
<p>Points: 351</p>
<p># Comments: 249</p>
]]></description><pubDate>Mon, 26 Jan 2026 15:38:50 +0000</pubDate><link>https://www.greptile.com/blog/ai-code-review-bubble</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46766961</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46766961</guid></item><item><title><![CDATA[Every GitHub object has two IDs]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.greptile.com/blog/github-ids">https://www.greptile.com/blog/github-ids</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46602591">https://news.ycombinator.com/item?id=46602591</a></p>
<p>Points: 327</p>
<p># Comments: 75</p>
]]></description><pubDate>Tue, 13 Jan 2026 15:52:33 +0000</pubDate><link>https://www.greptile.com/blog/github-ids</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46602591</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46602591</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>We have ways to approximate our impact on code quality, because we track:<p>- Change in number of revisions made between open and merge before vs. after greptile<p>- Percentage of greptile's PR comments that cause the developer to change the flagged lines<p>Assuming the author is will only change their PR for the better, this tells us if we're impacting quality.<p>We haven't yet found a way to measure absolute quality, beyond that.</p>
]]></description><pubDate>Wed, 17 Dec 2025 23:25:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=46307025</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46307025</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46307025</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>Apologies, that is poor wording on our part. It's internal data from engineers that use Greptile, which are tens of thousands of people from a variety of industries. As opposed to external, public data, which is where some of the charts are from.</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:46:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305980</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305980</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305980</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>Most of our customers are enterprises, so I feel relatively comfortable assuming they have some decent testing and QA in place. Perhaps I am too optimistic?</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:39:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305881</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305881</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305881</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>Thanks! The first 4 charts as well as Chart 2.3 are all from our data!</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:38:23 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305871</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305871</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305871</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>This is a good one, wish we had included it. I'd run some analysis on this a while ago and it was pretty interesting.<p>An interesting subtrend is that Devin and other full async agents write the highest proportion of code at the largest companies. Ticket-to-PR hasn't worked nearly as well for startups as it has for the F500.</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:26:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305731</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305731</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305731</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>How would you measure code quality? Would persistence be a good measure?</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:25:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305709</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305709</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305709</guid></item><item><title><![CDATA[New comment by dakshgupta in "The State of AI Coding Report 2025"]]></title><description><![CDATA[
<p>This is a great suggestion. I'll note it down for next years. Curious, do you think this would be a good proxy for code quality?</p>
]]></description><pubDate>Wed, 17 Dec 2025 21:24:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=46305699</link><dc:creator>dakshgupta</dc:creator><comments>https://news.ycombinator.com/item?id=46305699</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46305699</guid></item></channel></rss>