The biggest friction point for a new user is figuring out where to start. The product does a lot — QA, retrieval, orchestration, memory — but it's not immediately clear who the primary audience is or what the "first win" looks like.
A guided onboarding flow (step-by-step tooltips on first use) would help a lot. Libraries like Driver.js, Shepherd.js, or React Joyride make this straightforward to implement. A short 5-minute "get your first result" walkthrough would dramatically reduce the time to aha moment and clarify whether Clean is aimed at solo devs, team leads, or AI-power users — right now it feels like it could be all three, which makes the entry point confusing.
Cursor
Kilo Code
Claude Code
I’m particularly interested in seeing how the orchestrator handles multi-agent handoffs for larger codebases. The automated QA at scale also sounds like a dream for maintaining software integrity without the usual manual overhead.
Checking it out on right now.
Can't wait to see how the token efficiency holds up with complex retrievals!
@aadhyaa_gauli Thanks Aadhyaa, that’s exactly the area we’ve been obsessing over.
For larger codebases, we’re treating context and memory as first-class orchestration primitives, not just stuffing more into the prompt. Each agent gets scoped context relevant to its task, while the orchestrator maintains a higher-level understanding of the repo, prior decisions, handoffs, and long-term project memory.
The goal is to make agents useful across long-running engineering workflows: remembering architectural choices, avoiding repeated discovery, retrieving only the most relevant files/notes/history, and handing off work without losing intent.
Token efficiency is a huge part of that. We’re focused on keeping retrieval selective and relevance-driven so complex tasks don’t degrade into giant context dumps. Would love to hear how it holds up on your codebase once you try it.
The 'self-improving' aspect is what caught my eye here. Does Clean learn from the manual edits I make after it generates code? It would be a massive productivity boost if the model started picking up on my specific architectural patterns or 'house styles' the more I use it.
I'm a CS student currently building in the AI space and love seeing tools that solve the 'context-drift' problem. I even shared a quick post about what you guys are doing over on LinkedIn! Always open to collaborating on meaningful projects like this.
LinkedIn Post: https://www.linkedin.com/posts/sujal-kishore-kumar-talreja-65975b216_clean-the-first-self-improving-ide-that-share-7458938902700138496-JA1p?utm_source=share&utm_medium=member_desktop&rcm=ACoAADaSluUBOuckqBc1BiJG90rMyKi4JZ5s5vU
Portfolio: sktalreja.vercel.app
As a frontend dev, loved the sky inspired blues in the theme, and how this looks great even in light mode, while most of the sites today rather push for dark mode. Cool Work Guys
@shiva1803 Thank you so much!
LESGGOO GUYS
amazing work !!
@ygloria_gao Thank you!!
The idea of a self-improving IDE that learns team patterns is the real winner here. Most tools just give generic suggestions, but if this can actually adapt to our specific codebase architecture and style over time, it’s a massive unlock for onboarding new devs. Excited to see how the memory holds up as the codebase grows.. @pavankumarny