CodeYam

CodeYam is one of 39 startups building alongside Vercel over six weeks as part of the Vercel AI Accelerator. See you at demo day!

If you haven't checked out the launch yet and would like to, you can find it here: producthunt.com/products/codey…

CodeYam CLI & Memory is now live on Product Hunt! We are currently ranked #16 for Product Hunt launches today. Thank you to everyone who's supported our launch and shared reactions or feedback on what we've built.

More context: Background blog post: open.substack.com/pub/codeyam/p/… 90 sec demo on our own repo: youtu.be/oJ2gTb-lxbE Demo teaching Claude a real OSS repo (Plane): youtu.be/CjOKBwBCcOs Website: codeyam.com

How to Get Started: Install: npm install -g @codeyam/codeyam-cli@latest Then from your project root run: codeyam

Introducing CodeYam Memory – comprehensive memory management for Claude Code. We built CodeYam Memory because Claude Code kept making the same mistakes on our codebase. Our claude dot md files quickly got stale and maintaining by hand or with Claude wasn’t sufficient.

Our team's first Show HN launch is live 🚀 CodeYam CLI & CodeYam Memory for Claude Code, now available.

If you use Claude Code heavily, you’ve probably felt this: - You fix something once. - You explain a pattern once. - You clarify an architectural decision once. And then a few sessions later (or even later in the same session)… you’re explaining it again. We’ve been building something at CodeYam to solve exactly this problem. A memory layer for Claude Code that helps your agent: - Remember architectural decisions - Maintain useful rules and conventions across sessions - Stop re-investigating the same parts of your codebase - Avoid repeating the same misunderstandings It’s designed for builders who are already deep in agent-based workflows and want their AI to actually improve over time, not reset every session. We’re opening early access soon. If you’re a heavy Claude Code user and want to try it first, join the waitlist here: codeyam.com/waitlist

Full write-up here: blog.codeyam.com/p/how-were-usi…

Communicating a complex developer tool can be almost as challenging as building it. Our team has been using AI not just to write code, but to pressure test how the product is explained. Write-up by product designer Dani Raskovsky in the replies.

Tool use isn’t a sign of weaker intelligence. For complex software problems, it’s often the opposite. A framework for thinking about when AI should “just do it” vs when tools are essential: 👉 blog.codeyam.com/p/to-tool-or-n…

We asked 7 AI tools to build the same app from a single prompt. The results were typically fast, sometimes buggy, and consistently fascinating. 🚨 New on the CodeYam blog → blog.codeyam.com/p/creating-sof…

🚨 Devs, EMs, CTOs, founders, and software builders: We're running paid 1:1 research sessions on April 23 & 25 to get feedback on a new developer tool we’re building. ✅ $50 gift card ✅ 60 mins on Zoom ✅ Camera on, remote Interested? Fill this out 👉 forms.gle/rjUe3YzKxjkffa…

🤔 Curious about if the current wave of AI-native prototyping tools are as magical as early adopters say? Our team is too. Our product designer put @v0, @cursor_ai, @boltdotnew and @lovable to the test to redesign our website. Full comparison here: blog.codeyam.com/p/testing-the-…

We explore two approaches to tackling the AI communication challenge: 1. Iteration (costly but necessary) 2. Better tooling (reduces iteration costs) The right tools can make AI collaboration far more effective.

New blog post: "AI Can't Read Your Mind...But Iteration Is Expensive". No matter how advanced AI becomes, the communication gap in complex projects remains the fundamental challenge. Read our latest thoughts: blog.codeyam.com/p/ai-cant-read…