Bitcollage

@ilkerulusoy @fkadev 😂

@fkadev TUI <3

Let’s quit using TUIs.

@sasha_zelts @aidenybai That’s up to you. Just be careful with the licence terms. As an individual or small team, you can’t really afford MIT, but overly strict licence terms aren’t ideal either.

@Bitcollage @aidenybai Yeah, should I open ?

what if ghostty had vertical tabs? i'm too lazy to learn tmux and i want an interactive UI to manage my agents/terminals

@sasha_zelts @aidenybai closed source?

would be honoured to have you test and potentially collaborate on acepe.dev Aiden ✌🏼 it's in beta still but it has: 4 workings agents (Claude code, Codex, Cursor, Opencode) - unified skill management interface (sync skills accross the agents within the app - attention queue (5 state: Input needed, planning, working, error, finished) - side by side panel/full screen mode - plan mode integration for cursor, claude code, codex. right side panel spawns with beautifully markdown rendered plan. - SQL studio/s3 connector, see your dev/prod data directly from the app. - Model settings PER agent PER mode eg: opus 4.6 for plan, sonnet 4.6 for build mode for Claude Code - Terminals panels - Review panels like in cursor with undo/accept - Checkpoints feature that is AGENT AGNOSTIC, meaning we don't rely on claude code vibe coded inner system - File chips in the conversation with git diff and file panel on click. - PR and commit chips with PR and commit panels - git integration to create commits, see staged, unstaged files. - worktree integration that is agent agnostic. - unlimited number of projects in parallel - Todo list UI, AskUserQuestion UI, Permission UI. - virtualisation for the session content for performances. - @ for files and / for skills dropdowns with preview. • polished chips in the agent panel input for text copied, image s, file mentions. Unlike other apps, this took 2.5 months to build and was human reviewed. Ask me about an architecture or how some logic works and I will be able to answer you straight away. It is not currently not OSS, but maybe I should open source it after all

@juristr It was more like speed dating 😂 …I have my special criterias. Cmux was love at first sight. > ALL cmux feels and behaves more like a terminal.

@Bitcollage All?? 👀. Wow, respect 😅. I sticked to superset since over a week now with occasionally just falling back to regular tmux. that seems to work for me really well

Another one for the collection - cmux.dev - conductor.build - superset.sh - (what did I forget??) - dmux.ai

@youyuxi @voidzerodev Angular?

Introducing Void, the Vite-native deployment platform: 🚀 Full-stack SDK ⚙️ Auto-provisioned infra (db, kv, storage, AI, crons, queues...) 🔒 End-to-end type safety 🧩 React/Vue/Svelte/Solid + Vite meta-frameworks 🌐 SSR, SSG, ISR, islands + Markdown 🤖 AI-native tooling ☁️ One-command deploys void.cloud

@juristr @superset_sh @FlyaKiet Does logging in with an Anthropics account violate the Claude Code Subscription policies?

@Bitcollage @superset_sh yeah at the core it's a terminal wrapper. You're still using the terminal and whatever auth method you're using for that. They have a chat feature which I haven't been using for which you need to auth. Maybe @FlyaKiet has some more insights whether u can use ur max subs there

Said it before, @superset_sh hits the sweet spot for my workflow. Some team mates asked, so I quickly recorded this quick walkthrough. 👉 manages worktrees 👉 shows diff 👉 GH info (e.g. connected PRs, pipeline status) 👉 integrated terminal survives app restarts and more.

@juristr @superset_sh I see. You are using Max.

@juristr @superset_sh Works with the Claude code subscription?

@jim_stef @MrAhmadAwais It is officially, see x.com/addyosmani/sta…

@MrAhmadAwais > Disclaimer > This is not an officially supported Google product. stop spreading misinformation

Google finally launched an official CLI (gmail, drive, sheets, gmail, calendar, admin, keep, meet 10+ more). $ npm i -g @⁠googleworkspace/cli $ npx skills add github:googleworkspace/cli My prediction of 2026 being the year of CLI is coming true.

Responsibility is clearly shifting more towards the developer. LLMs and agents make this part of the work much easier. With good tests, skills, and guard rails, developers can ensure a high level of quality even before opening the PR. This also fits in with Simon Willisons Agentic Engineering Anti-Patterns. Developers should already be sure that the code works before submitting it for review. The reviewer is no longer the one who produces the quality, but the one who verifies it.

After preparing for exams for 3 months, clearing 2 exams, paying taxes for 4 years , I decided to apply for the German Permanent Residency, wish me luck 🤞. I hope the German bureaucracy is kind to me 🤗

@youyuxi ...and are they sponsoring Vite?

MacBook Pro M5 demo is building a Vite app.

I would not be surprised if @AnthropicAI made all of its models and weights open-source in an emergency.

The latest release of the analogjs/angular-skills repo has been updated according to the recommendations provided by @mgechev github.com/analogjs/angul…

realizing my software engineering is cooked in about 3 years and i need to figure out what to do next

@kentcdodds @karpathy same here!

@karpathy Agreed. The last two months have shifted my software development workflow significantly. I've closed every issue on my actively developed projects and added new features I never even bothered opening an issue on because I "knew" I'd never have the time to build them. I love it!

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but specifically this last December. There are a number of asterisks but imo coding agents basically didn’t work before December and basically work since - the models have significantly higher quality, long-term coherence and tenacity and they can power through large and long tasks, well past enough that it is extremely disruptive to the default programming workflow. Just to give an example, over the weekend I was building a local video analysis dashboard for the cameras of my home so I wrote: “Here is the local IP and username/password of my DGX Spark. Log in, set up ssh keys, set up vLLM, download and bench Qwen3-VL, set up a server endpoint to inference videos, a basic web ui dashboard, test everything, set it up with systemd, record memory notes for yourself and write up a markdown report for me”. The agent went off for ~30 minutes, ran into multiple issues, researched solutions online, resolved them one by one, wrote the code, tested it, debugged it, set up the services, and came back with the report and it was just done. I didn’t touch anything. All of this could easily have been a weekend project just 3 months ago but today it’s something you kick off and forget about for 30 minutes. As a result, programming is becoming unrecognizable. You’re not typing computer code into an editor like the way things were since computers were invented, that era is over. You're spinning up AI agents, giving them tasks *in English* and managing and reviewing their work in parallel. The biggest prize is in figuring out how you can keep ascending the layers of abstraction to set up long-running orchestrator Claws with all of the right tools, memory and instructions that productively manage multiple parallel Code instances for you. The leverage achievable via top tier "agentic engineering" feels very high right now. It’s not perfect, it needs high-level direction, judgement, taste, oversight, iteration and hints and ideas. It works a lot better in some scenarios than others (e.g. especially for tasks that are well-specified and where you can verify/test functionality). The key is to build intuition to decompose the task just right to hand off the parts that work and help out around the edges. But imo, this is nowhere near "business as usual" time in software.

@Nartc1410 No implicit imports? No new authoring format?

re:updated Angular roadmap github.com/angular/angula… No way AI experience is more important then DX. Am I really missing something? 🥹

@bcherny I am not seen it in my max plan!

Sonnet 4.6 is now live in Claude Code. It's cheaper than Opus 4.6 and nears Opus-level intelligence, and devs in early testing often preferred it to Opus 4.5. Now the default for Pro and Team plans.