chrisrickard

damn... cursor doesn't get as much love as claude code or codex (on here anyways), but honestly, it's gotten amazing in the last 6 months.

i loved switching between sol / fable / grok 4.5.

... and something tells me fable might not be available on cursor for much longer :/

much needed, will give it a crack - congrats on shipping!

Of course it’s an Australian spider

Cursor 3 is a full rewrite. No VS Code

I’ve found it interesting too.

It’s like I spent 20 years mastering painting in watercolour… nerding out on other painters, canvas options, even the backstory on some guy that makes a specific paint etc.

And I don’t regret any of that. but now I’m just loving creating my art 100x faster.

I thought I loved the craft (and I did) but more, I loved the product.

… i’m still seeing a therapist about this time period.

Thank you, a refreshingly interesting read

I have read all your yearly reviews and you’ve been an inspiration Michael. Keep it up, and I’m glad your 8th year was your happiest yet!

Skills are chainable e.g. skill A can invoke skill B and then decide to invoke skill C etc… I don’t believe your slash commands can do this?

Very cool, I didn’t realise the web audio API was so extensive these days

I did this recently. Created a Skill that had access to executing very specific ific (reviewed) script for DB interaction, that connects to your a replica/anonymised DB, read only user, via VPN, via a jumpbox.

Really fun idea treating an album like a dependency!

Also, probably a good way to give LLMs more context compared to “div soup”

Amazing achievement guys, seriously impressive. Now onwards and upwards!

very cool concept. and great tracks. congrats on shipping!

Another 15-yearer here too! Thank you HN, and for all the work you do @dang and @tomhow

Because the ridiculous scope creep perhaps? And spending $96M of government money on an website (still with large faults that were backed out)

This was Accenture and Deloitte - not some backyard dev shop.

I’m working on Userdoc, a spec-driven development workspace.

Break down your software requirements (Userdoc guides you through the process), refine/confirm, setup your technical specs, coding/business guidelines & guardrails, and then create development plans (specs) which can be easily consumed by coding agents via MCP, or by platforms like Lovable / v0 using Markdown. Working on Cursor background agent integration atm.

https://userdoc.com

Your second point “planning in advance” could be referred to as spec-driven development… it’s a funny term in a sense (didn’t we always do that?), but I think your 7th point drives it home “a very weird form of management” - clear instructions, necessary context, and actionable feedback. As far as written words go, much more like waterfall than agile.

This is why AI development will be huge in the “Buy vs Build” space… Businesses (with a capable tech team)can build the 20% of the SaaS they need, and stop paying for the 80% every single month.

We are tackling this at https://userdoc.fyi - we help you build your specs (epics, stories, acceptance criteria, tech notes, test cases, etc) - then you can generate what we call Dev Plans, one or more requirement layers for implementation.

e.g maybe a dev plan is all your authentication feature requirements, or in the house of analogy – all the requirements for the rooms, but with instructions to actually just first build the floor, and the walls.

Dev plans then slice the reqs into meaningful units of work, as mentioned in the article – a feature/story, is often too large of a checkpoint, or often needs to be implemented in collaboration with other features/stories, so it understands the correct architectural context,.

You can then implement Dev plans over MCP, or copy to .md for tools like Lovable or V0.

Tools like Userdoc (https://userdoc.fyi) help in a few ways, you can easily create requirements (stories, personas, journeys, test cases), but also reverse engineer existing source code into detailed docs, then ask natural language questions etc. AI helps us plan our product in Userdoc, and our devs connect via MCP to bring those requirements directly in to Cursor (full disclaimer, I work at Userdoc - but we eat our own dogfood)

Vibe coding in the traditional sense (coined by Karpathy back in Feb): 20%

Vibe coding using detailed, structured requirements (from tools like Userdoc): 65%

Userdoc is primarily around building and maintaining software requirements, so we don't do competitor research, or watch market trends etc.

But we can import any info you already have (documents etc), and have an interactive wizard to break it down into features (stories, epics, NFRs, test cases, personas, journeys) then create detailed acceptance criteria for each feature.

We can also take existing source code, and reverse engineer that into detailed documentation.

The AI coding wars heat up…

I think at a high level you’re correct.

But no matter how much “vibe” is in the code, businesses still need to create, validate, review, and approve software requirements.

This is what we help with at Userdoc.fyi - AI augmented requirements, where humans are the input and the glue, but AI does the heavy lifting.

You then integrate the requirements into Cursor, Copilot, Aider, whatever - and then use vibe coding to build POC, and perhaps vibe engineering to take it to production.

This is Userdoc, AI augmented software requirements.

Similar to a feature we offer in Userdoc, we can reverse engineer a codebase not themes, features (stories), and highly detailed descriptions. In terms of of how useful it is, we’ve had many customers fing it extremely valuable… something that would have taken a human months now takes minutes.

Check out http://userdoc.fyi - helps build out super detailed requirements, user personas, and user journeys with AI

Userdoc has versioning of stories and acceptance criteria, aimed to at helping this exact issue https://userdoc.fyi (transparency: I’m the founder)