<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: caijia</title><link>https://news.ycombinator.com/user?id=caijia</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 10 Jun 2026 07:28:22 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=caijia" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by caijia in "Copilot edited an ad into my PR"]]></title><description><![CDATA[
<p>I've already be patient when claude code always signs my commits as co-author by defualt. Yes, it is.<p>But I'm also paying the plan. Theres something odd about a tool which i paid for using my output to AD itself.</p>
]]></description><pubDate>Mon, 30 Mar 2026 08:38:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=47571919</link><dc:creator>caijia</dc:creator><comments>https://news.ycombinator.com/item?id=47571919</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47571919</guid></item><item><title><![CDATA[New comment by caijia in "15 years, one server, 8GB RAM and 500k users – how Webminal refuses to die"]]></title><description><![CDATA[
<p>UML is a smart call, and reminds me when I built an inventory and shift scheduling system on wordpress in 2017.<p>somtimes the "wrong" / "old" tool for some job is exactly right for you if you really understand it. UML is old but fits here.<p>15 years is long enough to call memory about a lot of things.</p>
]]></description><pubDate>Mon, 30 Mar 2026 08:31:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=47571871</link><dc:creator>caijia</dc:creator><comments>https://news.ycombinator.com/item?id=47571871</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47571871</guid></item><item><title><![CDATA[New comment by caijia in "Lat.md: Agent Lattice: a knowledge graph for your codebase, written in Markdown"]]></title><description><![CDATA[
<p>I've been doing similar work since Claude code updated their "slash command"(later merge to skills), first 3-4 long content docs, gradually split to modular groups. I designed it for loading the docs based on what agent is actually doing. The maintenance  part is honestly not that hard, for me, I created some CI jobs that diffs the docs against the codebase and flags drift handles of it.<p>The pattern works.<p>But I keep catching myself spending more time on how to organize context  than on what the agent is actually supposed to accomplish.<p>Feels like the whole space has that problem right now.</p>
]]></description><pubDate>Sun, 29 Mar 2026 15:10:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=47563840</link><dc:creator>caijia</dc:creator><comments>https://news.ycombinator.com/item?id=47563840</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47563840</guid></item></channel></rss>