
Alrighty folks, I've renamed my twitter account from @zephraph to @just_be_dev. Got a placeholder for the old one.

Open-source enthusiast; maker. Co-host of @DevtoolsFM. @recursecenter Alum. Prev at @ValDotTown, @oxidecomputer, @ArtsyOpenSource.

Alrighty folks, I've renamed my twitter account from @zephraph to @just_be_dev. Got a placeholder for the old one.

This is a brilliant idea. Thinking through it, makes total sense that the agent using the MCP, who is experiencing frustration in the shape of the tools to achieve it's specific goal at the time, is perfectly suited to provide that feedback to you. And if the tool definition of the submit_feedback tool is good enough, it's just going to...magically happen. There are so many yet-to-be-discovered patterns like this that can help close the loop and hill climb for you within product development.
Lot of people asked about this. You can basically make your MCP self improving by doing this: - Add a submit_feedback tool to your MCP - Agent using your MCP will write detailed issue including proposed fix. - Have it send to an internal Slack channel - Have Claude in Slack channel and tag it to put up a PR when you agree w proposed change.

State of the tabletop: - Lessons from doing a bunch of collages - Images autolinked to video timestamps and URLs feels promising - Balancing snapping alignment vs freeform - Touch troubles

Shipped @nkzw/oxlint 3.0: Optimized for reducing agent mistakes in JS. * Enabled type-aware lint rules with tsrs * Adopted anti-slop rules from @dillon_mulroy, made them faster, renamed to "quality" * Enabled a few rules that @sebmck uses * Moved sorting into a separate module

Feeds may be the right interface for working with AI agents. Chatbots feel too heavy. Messages lack clear task boundaries. Feeds are lightweight, task-scoped, proactive but not stressful. So I built Aware Feed into our Aware browser—a personal feed where I can interact with all my AI agents in one place.


Lately I've been using rectangleapp.com/ for window management in OSX. What are y'all using?

brand explorations and style studies for @LocalCanApp Access

ur cad, ur drawings, and an ai agent that actually edits them. all in one app

exploring a visual language for the brand new @polar_sh usage billing engine ridiculously high-performant event ingestion, folds over the event stream & zero latency credit checks


Sneak peek of next blog! It has a reference to Nolan's Memento which I've always seen as a great analogy for working with agent memory

Introducing iOS for the web. This is Telegram (written in Swift and Objective-C), running fully in a browser on our custom renderer. No emulator, No iPhone required. You can infinitely branch the entire codebase by copy pasting a frame and make changes to each frame. We built a compiler+renderer (in Rust) that allows you to easily design on top of a native iOS codebase. Most edits land in under a second and is much much faster then xcode builds. Our goal is to take any codebase and turn it into a design environment that fully runs in your browser!

one of the most brilliant takes i've seen as of late on HCI in the era of LLMs Attention steering by Sun Chuanqi. Actually operating an LLM's attention and token generation. the concept, UX and sound design are just amazing. There are still so many unexplored patterns. this one's particularly inspiring github.com/chuanqisun/spaces…

Val Town's beautiful office is up for grabs! - Location: downtown brooklyn (325 gold st) - Rent: ~$9k/month all-in - Size: 2k sq ft - Available furnished or unfurnished - Very nice landlord We've loved it here, and would love the office to go to a friend. DM me :)

Updates to my anti-slop repo: 1. More @EffectTS_ rules 2. Vendored oxlint jev plugin, write rules as natural language questions for @typesafeai’s Jev or other decision models 3. Pre-baked effect + regular ts oxlint jev rules for (2) 4. Support for changed-code-only semantic rules powered by LLMs or decision models 5. Pre-baked rules for @martinfowler code smells that instruct your agent which refactorings to apply for (3) and (4) Https://github.com/k-mistele… We are in a (now slightly less) asymmetric war on slop per @swyx