Massively Parallel Procrastination

Simon Willison and I have been talking about hosting a small-group event (or maybe an event series) about agentic engineering for ages.

I'm pleased to announce the release of Superpowers 6.4.

A while back, Claude Opus built a pretty ambitious compiler.

You can also read this post on Prime Radiant's blog.

You can also read this post on Prime Radiant's blog.

As I've been spending more and more time with agentic development, it's more and more important to me that sessions run somewhere other than my laptop.

You can also read this post on our corporate blog at https://primeradiant.com/blog TL;DR: Superpowers 6 is much, much faster and burns many fewer tokens to get the same high-quality outcomes.

We're hiring someone to help out full time with Superpowers community and code work.

This afternoon, I pulled some data from backups of my .claude directory to build a visualization of Claude Code plugin installs over time.

I'm pleased to announce Superpowers 5.1.0.

Work gets better when it's reviewed after completion.

Last fall I had a bit of a problem with Claude.

Crossposted from Prime Radiant's blog – I'm really excited about all of the stuff we are doing at Prime Radiant.

One of the first things I stumbled into as I was building what became Superpowers was the idea of a "gate" in a prompt.

Over the past couple of months, Superpowers has gotten popular.

One of the things that keeps coming up in my conversations about AI and software is the difference between software with an agent or LLM in the loop and what we've always thought of as "software".

Superpowers 5 is out today.

Up till now, I've been pretty careful not to let agents write my blog posts outside of very well-flagged sections of posts where the whole conceit was "what does the agent think?"

I've realized that I have two primary ways that I'm building software with AI.

Last night, I had dinner with a friend from college.

When cryptocurrencies first became a thing, some of their loudest proponents talked about how they were free, democratic, and open, and didn't need all of the regulation and bureaucratic institutions that "fiat currency" needed.

A lot of the focus on interacting with agents recently has been on "context engineering," which is the discipline of packing exactly the information that the agent needs to know into the prompt.

] If you've been following consumer uptake of AI agents over the past few weeks, it will have been hard to miss Clawdbot (now known as Openclaw).

I woke up this morning to discover that, without my knowledge or consent, someone had created an ICO using the Superpowers name and my Twitter username.

This is just a quick post to say that I've started a company.

I've been using Linear as the project and issue tracking tool on a new project.

My all-time favorite word game is called 'Wordiest'.

I've been working through some testing of OpenAI's new skills support in their codex CLI tool.

I'm pleased to announce Superpowers 4.0.

I've heard from a number of folks that they're seeing Claude Code just...not use skills they have installed.

TL;DR: I built a tool to make it easier to run your favorite coding agent in a container without a lot of setup.

I've spent a fair bit of time over the past year having agents build webapps for me.

Quoting Claude Opus 4.5, whom I had just asked to read my post about MCP design: The philosophical shift: Design for the undertrained NOC kid at 2AM.

Superpowers is my software development workflow tool set for coding agents.

This week, I visited my friends at 2389 in Chicago.

I've been posting rather a lot about Anthropic's new skills system and about my Superpowers system that builds software development and skills development workflows using those building blocks.

I've got more that I should write about prompting for code reviews, but this simple prompt (for Claude Code) is way more effective than it has any right to be.

I'm pretty leery about "AI Browsers" for all the reasons that Simon is concerned about them The prompt injection risks are real.

Today, Anthropic announced that they're expanding availability for Claude.ai's memory feature to Pro and Max users.

I've been working on some new skills that might be useful for folks using my Superpowers skills and workflows in Claude Code, but I'm not 100% sure they're "right" yet, so I'm creating a new Claude Code plugin as a skills incubator.

TL;DR: I built a lightweight Chrome MCP.

Anthropic is releasing their first-party Skills system across Claude Code, Claude.ai and the Claude API, all launching today.

"It's not this — it's that."

Yesterday, I released an update to Superpowers.

I kind of despise agents thinking and talking about me as "the user", so I typically tell them to always think of me and address me as "Jesse" and that I'm their boss, but that we work together as partners. - That worked pretty well until I packaged up a…

It feels like it was just a couple days ago that I wrote up "How I'm using coding agents in September, 2025".

[Eagle-eyed readers will note that, as I write this, it's October 2025.

This is a very, very informal interim research report about something I've been playing with over the past couple days.

It looks like Moonshot AI have an Anthropic-compatible API endpoint for their new open frontier model, K2.

I've been playing around with the idea of building my own commandline coding agent, both to better understand how these things work and to try out a couple of things that I haven't seen in any of the existing tools.