Skill eval loop: 1. Pick a skill from your library 2. Write 10 test inputs 3. Run. See pass/fail per input. 4. Edit the skill. Run again. Compare scores. 5. Keep the version that scores higher. Built this into Skill Creator Desktop.

Today we are launching the Kapso CLI: WhatsApp numbers for agents. 1️⃣ npm install -g @kapso/cli 2️⃣ kapso setup Done, your agent has a WhatsApp number.

Last week our vibe-coded agent-native document editor, Proof, went viral. 4,000+ documents were created in the first two days—and then the app started crashing uncontrollably. I vibe coded all of Proof—so I spent the next week or so not sleeping and watching Codex agents debug

200+ SKILL.md files across 6 projects. The management tool was grep and find. Needed: - Library view with search, tags, filters - Version diffs when editing a skill - Test runner that checks pass rates after changes - One place instead of 6 repos So I built it. Desktop app,

Agent runtimes got 10x better in the last year. Skill tooling is still "copy this Markdown file into your project." No package manager conventions, no testing standard, no versioning. We solved this for code decades ago.

You can now schedule recurring cloud-based tasks on Claude Code. Set a repo (or repos), a schedule, and a prompt. Claude runs it via cloud infra on your schedule, so you don’t need to keep Claude Code running on your local machine.

to my fellow "agentic engineer" who opens Finder to find what Claude just edited learn some vim https://t.co/hVaXK6FlP3 put together a straight forward simple tutorial of vim [use lazyvim] learn 6 vim commands. never touch a mouse again.

I've been an engineer for nearly a decade. Right now, process has never been more important. And skills are the best way to bundle up processes for agents. Here are the 5 I use every day: /grill-me /write-a-prd /prd-to-issues /tdd /improve-my-codebase

Turned my blog articles into one big design engineering skill that you can use with coding agents like Claude Code or Codex. It covers animations, component design, principles from my open source projects like Sonner, and more. https://t.co/WKXPGQuvQd

Introducing napkin 🧻 (Agents + Obsidian + a CLI tailored for agents is amazing). napkin is a knowledge base that doubles as an agent memory layer. ultra fast, completely sidestepping vector search. First class support for @badlogicgames's pi - includes a self-distillation

MCP server trained on 50k tweets and 500 images. claude mcp add vvriter -- npx vvriter https://t.co/PDiLq6rvFF

If your company isn't building internal agent skills, you're falling behind the gap it creates in terms of efficiency is as much as companies using vs not using ai Every workflow your team runs repeatedly should be a skill. Every internal API should have an agent definition.

We doubled Claude usage on weekends, and outside 5–11am PT on weekdays for the next 2 weeks.

https://t.co/BBIva1uPHn

OK, well. I ran /autoresearch on the the liquid codebase. 53% faster combined parse+render time, 61% fewer object allocations. This is probably somewhat overfit, but there are absolutely amazing ideas in this.

Breaking: Proof! We’re temporarily down due to heavy load. Remind me never to launch a new product and then have a flight early the next morning Was literally standing next to TSA with my laptop open trying to fix the outage. Now on plane, about to take off! Will start

You've built 100+ agent skills. Now what? No way to evaluate which ones actually work. No version tracking when you iterate. No live testing environment. Just a pile of Markdown files scattered across projects. Which ones are bloat, whats actually useful, what can be merged

TLDR: it is a cron job dispatching tickets from Linear to workers, each of which is a Ralph loop using a Linear comment as draft pad for persisted state. Yes it is all you need. Beautifully designed and minimal. https://t.co/g05ImsJIZh

very cool project, may our vibe coding be our Ibadah

OpenAI published Symphony — a spec for turning project work into autonomous implementation runs. Manage work, not agents. I took that idea and rebuilt it with Claude Code @AnthropicAI hatice is my implementation of the Symphony spec, powered by Claude Code Agent SDK. It