<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: itsderek23</title><link>https://news.ycombinator.com/user?id=itsderek23</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sun, 12 Apr 2026 11:42:42 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=itsderek23" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by itsderek23 in "We've raised $17M to build what comes after Git"]]></title><description><![CDATA[
<p>How I'm using git/Github has changed with agentic coding. However, I'm not using swarms of agents to write code, so it's bit hard for me to decipher the JTBD of gitbutler.<p>Another take I've seen is <a href="https://agentrepo.com/" rel="nofollow">https://agentrepo.com/</a>, which is light-weighted hosted git that's easy for agents to use (no accounts, no API keys, public repos are free). There are large parts of the GitHub experience I'm no longer using (mostly driving from Claude), so I think this is an interesting take.</p>
]]></description><pubDate>Fri, 10 Apr 2026 15:01:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=47719195</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=47719195</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47719195</guid></item><item><title><![CDATA[Show HN: Subtle – Local, open-source analytics for Claude Code sessions]]></title><description><![CDATA[
<p>I’m used to having good observability for systems and workflows, but now that I generally start development with Claude Code vs my IDE, I lost a layer of visibility into how my code is produced.<p>Subtle (<a href="https://github.com/itsderek23/subtle" rel="nofollow">https://github.com/itsderek23/subtle</a>) is a free, open-source app for analyzing Claude Code logs that runs entirely on your local machine. It lets you:<p>* See overall Claude Code usage over time (broken down by AI vs tool time)
* Count Claude-assisted Git commits over time
* Visualize individual sessions with a trace showing how time is spent
* Filter sessions by message text
* View session transcripts<p>I don’t yet know what “good” Claude Code usage looks like ... fewer interventions? Smaller commits and PRs? Lower code churn within a session? Would love your thoughts.</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46590647">https://news.ycombinator.com/item?id=46590647</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 12 Jan 2026 16:26:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=46590647</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=46590647</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46590647</guid></item><item><title><![CDATA[Streamlining the eval of LLM code gen / eval agents]]></title><description><![CDATA[
<p>Article URL: <a href="https://dlite.cc/2023/10/04/2023-eval-rag-apps.html">https://dlite.cc/2023/10/04/2023-eval-rag-apps.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=37956835">https://news.ycombinator.com/item?id=37956835</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Fri, 20 Oct 2023 14:45:53 +0000</pubDate><link>https://dlite.cc/2023/10/04/2023-eval-rag-apps.html</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=37956835</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37956835</guid></item><item><title><![CDATA[DevOps AI Assistant Open Leaderboard]]></title><description><![CDATA[
<p>Article URL: <a href="https://github.com/opstower-ai/devops-ai-open-leaderboard">https://github.com/opstower-ai/devops-ai-open-leaderboard</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=37600688">https://news.ycombinator.com/item?id=37600688</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 21 Sep 2023 17:11:16 +0000</pubDate><link>https://github.com/opstower-ai/devops-ai-open-leaderboard</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=37600688</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37600688</guid></item><item><title><![CDATA[New comment by itsderek23 in "Show HN: Release AI – Talk to Your Infrastructure"]]></title><description><![CDATA[
<p>Nice work! I'm working on a similar standalone DevOps AI Agent (OpsTower.ai). This post shows how the agent is structured and how it performs against a 40 question evaluation dataset: <a href="https://www.opstower.ai/2023-evaluating-ai-agents/" rel="nofollow noreferrer">https://www.opstower.ai/2023-evaluating-ai-agents/</a></p>
]]></description><pubDate>Fri, 25 Aug 2023 18:31:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=37265435</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=37265435</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37265435</guid></item><item><title><![CDATA[Beyond demoware: how do you evaluate an AI agent?]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.opstower.ai/2023-evaluating-ai-agents/">https://www.opstower.ai/2023-evaluating-ai-agents/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=36971304">https://news.ycombinator.com/item?id=36971304</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 02 Aug 2023 14:22:32 +0000</pubDate><link>https://www.opstower.ai/2023-evaluating-ai-agents/</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=36971304</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36971304</guid></item><item><title><![CDATA[New comment by itsderek23 in "Things I wish I knew before building a GPT agent for log analysis"]]></title><description><![CDATA[
<p>Thanks!<p>> Did you hit token limits?<p>While i used TikToken to limit the message history (and keep below the token limit), generally I found that I didn't get better completions by putting a lot of data into the context. Usually the completions got more confusing. I put a limited amount of info into the context and have generally stayed below the token limit.<p>>  Are you storing message/ chat histories between sessions<p>Right now, yes. It's pretty important to store everything (each request / response) to debug issues with prompt, context, and the agent call loop.</p>
]]></description><pubDate>Tue, 06 Jun 2023 16:41:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=36215494</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=36215494</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36215494</guid></item><item><title><![CDATA[Things I wish I knew before building a GPT agent for log analysis]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.logpal.ai/2023/06/01/2023-gpt4-api-gotchas.html">https://www.logpal.ai/2023/06/01/2023-gpt4-api-gotchas.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=36214970">https://news.ycombinator.com/item?id=36214970</a></p>
<p>Points: 3</p>
<p># Comments: 2</p>
]]></description><pubDate>Tue, 06 Jun 2023 16:10:56 +0000</pubDate><link>https://www.logpal.ai/2023/06/01/2023-gpt4-api-gotchas.html</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=36214970</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36214970</guid></item><item><title><![CDATA[Using Ruby to query a Rails DB with natural language]]></title><description><![CDATA[
<p>Article URL: <a href="https://dlite.cc/2023/05/06/2023-boxcars-ruby-intro.html">https://dlite.cc/2023/05/06/2023-boxcars-ruby-intro.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=36072533">https://news.ycombinator.com/item?id=36072533</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 25 May 2023 16:06:29 +0000</pubDate><link>https://dlite.cc/2023/05/06/2023-boxcars-ruby-intro.html</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=36072533</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36072533</guid></item><item><title><![CDATA[Novi battery maker sends Tesla S to the U.P. and back on one charge]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.freep.com/story/money/business/columnists/carol-cain/2022/01/05/tesla-model-s-our-next-energy-gemini-battery/9101549002/">https://www.freep.com/story/money/business/columnists/carol-cain/2022/01/05/tesla-model-s-our-next-energy-gemini-battery/9101549002/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=29816488">https://news.ycombinator.com/item?id=29816488</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 05 Jan 2022 23:07:10 +0000</pubDate><link>https://www.freep.com/story/money/business/columnists/carol-cain/2022/01/05/tesla-model-s-our-next-energy-gemini-battery/9101549002/</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=29816488</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=29816488</guid></item><item><title><![CDATA[New comment by itsderek23 in "I sold Baremetrics"]]></title><description><![CDATA[
<p>+1. Jonathan is great to work with if you are in a similar position as Baremetrics.</p>
]]></description><pubDate>Tue, 10 Nov 2020 21:18:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=25052063</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=25052063</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=25052063</guid></item><item><title><![CDATA[GitHub Codespaces for Machine Learning]]></title><description><![CDATA[
<p>Article URL: <a href="https://dlite.cc/2020/05/26/github-codespaces-machine-learning.html">https://dlite.cc/2020/05/26/github-codespaces-machine-learning.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=23324054">https://news.ycombinator.com/item?id=23324054</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 27 May 2020 14:41:53 +0000</pubDate><link>https://dlite.cc/2020/05/26/github-codespaces-machine-learning.html</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=23324054</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=23324054</guid></item><item><title><![CDATA[Machine learning deserves its own flavor of Continuous Delivery]]></title><description><![CDATA[
<p>Article URL: <a href="https://towardsdatascience.com/machine-learning-deserves-its-own-flavor-of-continuous-delivery-f4d1e76e6b69">https://towardsdatascience.com/machine-learning-deserves-its-own-flavor-of-continuous-delivery-f4d1e76e6b69</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22961053">https://news.ycombinator.com/item?id=22961053</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 23 Apr 2020 20:38:57 +0000</pubDate><link>https://towardsdatascience.com/machine-learning-deserves-its-own-flavor-of-continuous-delivery-f4d1e76e6b69</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22961053</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22961053</guid></item><item><title><![CDATA[Machine learning deserves its own flavor of Continuous Delivery]]></title><description><![CDATA[
<p>Article URL: <a href="https://booklet.ai/blog/continuous-delivery-machine-learning-cd4ml/">https://booklet.ai/blog/continuous-delivery-machine-learning-cd4ml/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22899347">https://news.ycombinator.com/item?id=22899347</a></p>
<p>Points: 5</p>
<p># Comments: 0</p>
]]></description><pubDate>Fri, 17 Apr 2020 14:42:27 +0000</pubDate><link>https://booklet.ai/blog/continuous-delivery-machine-learning-cd4ml/</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22899347</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22899347</guid></item><item><title><![CDATA[New comment by itsderek23 in "Show HN: Cortex – Open-source alternative to SageMaker for model serving"]]></title><description><![CDATA[
<p>Great Caleb - makes sense. Thanks!</p>
]]></description><pubDate>Tue, 14 Apr 2020 21:50:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=22871925</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22871925</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22871925</guid></item><item><title><![CDATA[New comment by itsderek23 in "Show HN: Cortex – Open-source alternative to SageMaker for model serving"]]></title><description><![CDATA[
<p>This certainly looks like a cleaner way to deploy an ML model than SageMaker. Couple of questions:<p>* Is this really for more intensive model inference applications that need a cluster? It feels like for a lot of my models, a cluster is overkill.<p>* A lot of the ML deployment (Cortex, SageMaker, etc) don't see to rely on first pushing changes to version control, then deploying from there. Is there any reason for this? I can't come up for a reason why this shouldn't be the default. For example, this is how Heroku works for web apps (and this is a web app at the end of the day).</p>
]]></description><pubDate>Tue, 14 Apr 2020 20:31:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=22871080</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22871080</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22871080</guid></item><item><title><![CDATA[Make ML Model Deplyment less painful: Setup Sagemaker Local Mode and PyTorch]]></title><description><![CDATA[
<p>Article URL: <a href="https://booklet.ai/blog/aws-sagemaker-pytorch-local-dev-flow/">https://booklet.ai/blog/aws-sagemaker-pytorch-local-dev-flow/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22794490">https://news.ycombinator.com/item?id=22794490</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 06 Apr 2020 15:19:21 +0000</pubDate><link>https://booklet.ai/blog/aws-sagemaker-pytorch-local-dev-flow/</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22794490</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22794490</guid></item><item><title><![CDATA[You may not need Airflow – yet]]></title><description><![CDATA[
<p>Article URL: <a href="https://towardsdatascience.com/you-may-not-need-airflow-yet-a95fec16f07">https://towardsdatascience.com/you-may-not-need-airflow-yet-a95fec16f07</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22742204">https://news.ycombinator.com/item?id=22742204</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 31 Mar 2020 20:31:36 +0000</pubDate><link>https://towardsdatascience.com/you-may-not-need-airflow-yet-a95fec16f07</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22742204</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22742204</guid></item><item><title><![CDATA[An End-to-End Lead Scoring ML Model with MLFlow+Sagemaker+Booklet.ai+Intercom]]></title><description><![CDATA[
<p>Article URL: <a href="https://towardsdatascience.com/a-true-end-to-end-ml-example-lead-scoring-f5b52e9a3c80">https://towardsdatascience.com/a-true-end-to-end-ml-example-lead-scoring-f5b52e9a3c80</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22683919">https://news.ycombinator.com/item?id=22683919</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 25 Mar 2020 12:33:58 +0000</pubDate><link>https://towardsdatascience.com/a-true-end-to-end-ml-example-lead-scoring-f5b52e9a3c80</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22683919</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22683919</guid></item><item><title><![CDATA[Covid-19 International Social Science Research Tracker]]></title><description><![CDATA[
<p>Article URL: <a href="https://github.com/natematias/covid-19-social-science-research">https://github.com/natematias/covid-19-social-science-research</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22637141">https://news.ycombinator.com/item?id=22637141</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Fri, 20 Mar 2020 12:52:19 +0000</pubDate><link>https://github.com/natematias/covid-19-social-science-research</link><dc:creator>itsderek23</dc:creator><comments>https://news.ycombinator.com/item?id=22637141</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22637141</guid></item></channel></rss>