XRelevance 90
“You didn't read the thing when you generated it, I won't read it when I'm reading it.”
“You didn't read the thing when you generated it, I won't read it when I'm reading it.”
Open sourcehnRelevance 90
Show HN: A local merge queue for parallel Claude Code agents
The local merge queue for parallel Claude Code agents - funador/claude-code-merge-queue Claude Code Merge Queue 🚦 The local, zero-cost merge queue for parallel Claude Code agents. Several agents land, build, and test at the same time — this serializes it so pu...
Open sourceRedditRelevance 90
AI coding tools are getting good enough to actually ship things, which is kind of a problem for learning
Been tinkering with a SaaS side project for a few months and the gap between what I can ship now versus a year ago is genuinely strange. Not in a purely good way either. The tools are good enough that I can move fast through parts of the stack I barely underst...
Open sourceRedditRelevance 90
What do you actually do with your AI agents once they're finished?
I've been following the AI agent space for a while now, and there's one thing I can't seem to figure out. Building AI agents seems to be getting easier every month, but I rarely see people talking about what happens after they're built. If you've created an AI...
Open sourceHF Daily PapersRelevance 90
Agent Retrieval Bench: Evaluating Repository Context Retrieval for Coding Agents
Modern coding agents are usually evaluated by whether they eventually produce a correct patch, but patch generation depends on an earlier context-acquisition stage: finding the repository files needed for the task. We introduce Agent Retrieval Bench, a file-le...
Open sourceRedditRelevance 90
When does an AI-built app stop being a prototype?
I’m starting to worry that the AI app we’re building is basically a really impressive demo, not an actual product yet. We used agents to move fast and it worked almost too well. The UI came together, flows look good, customers can click through it, and everyon...
Open sourceRedditRelevance 90
i will not promote: I keep building useful things with Codex and Claude Code, but sharing them is still a mess
I keep building small things with Codex and Claude Code: expense trackers, dashboards, automations, internal tools, and random apps that solve problems I have. Some of them could genuinely be useful to other people. But turning every small build into a proper ...
Open sourcehnRelevance 90
Engineering management after the cost of code collapsed
I have been a director of engineering for a bit over three years now, and I still hear and read what I call the "old rules" repeated over and over: a director should not spend time coding, good work takes time, protect the team from the business, get consensus...
Open sourceRedditRelevance 90
ChatGPT read our support calls and now I owe my customers an apology
We've run NPS and customer satisfaction surveys for years and they always come back fine (mostly 8s and 9s with a few kind words). But our churn never matched how happy those numbers said everyone was, so at some point i took a big batch of our real sales and ...
Open sourceRedditRelevance 90
We spent $801 in AI coding credits in one month. Here is where it created value and where it created waste.
Last month, we spent $801.94 across 53 AI coding sessions on one product. The product is Mune , a privacy-first finance app with: on-device AI Story Insights Safe to Spend and budgeting subscription tracking Money Flow shared expense groups and Split AI create...
Open sourceRedditRelevance 90
I automated myself out of customer support and my SaaS quietly got worse. The rehumanising story.
Support was eating my time, so I did what everyone says to do, automated it, deflection flows, canned responses, a bot for the front line. My time freed up and I felt efficient. Then I noticed the slow rot: churn crept up, my reviews got cooler, and I'd lost t...
Open sourceRedditRelevance 90
How much of what you generate, actually makes it out of the door?
I did something slightly depressing on sunday. i went through my last month of ai outputs, all of it, drafts and images and scripts and little snippets, and counted how many actually got used somewhere real. published, sent, shipped, shown to a client. the num...
Open source