Now

What I’m Focusing On Now#

Professional Life#

• SVP, Engineering at Lark, March 2023 - May 2025.
• VP Engineering at Strive Health. Joined October 2025.
• Chief Tentacle Officer, Groktopus LLC. Still alive, still writing about AI and the future of work over at groktop.us.

What Changed Since I Last Updated This Page#

August 25, 2026#

A router for the home lab. I finally got around to one of my home-lab to-dos: setting up LiteLLM as a routing layer between my different harnesses and the inference providers behind them. It took a little work, but it was worth it. Now I can change my preferred models in one place, and all of my harnesses pick up the change without any extra work on my part. It’s a pretty cool open-source project, and worth checking out.

August 23, 2026#

Social media has been a bust. I’ve spent too much time on Twitter lately. Yes, I know we’re supposed to call it X now. I’m going to keep saying Twitter, because calling it X makes it sound like a porn site, and that’s not what we’re talking about. The place has really gone downhill. It always had a dark side, but ever since Elon Musk took over, the dark side runs the place. People openly call each other some of the worst racial epithets out there, they got rid of the moderators, so there are no consequences, and people are just as mean as they want to be. I find it really toxic. The only reason I’m still there is that Twitter is the epicenter of social media activity for the AI community, and I don’t know anymore if it’s worth it.

Every alternative has its own catch. The people who care most about inclusivity and codes of conduct tend to hang out on Bluesky, but Bluesky has a strong anti-AI mob mentality, so there’s really no AI community to speak of there, at least none I’ve found. I run my own Fediverse instance on GoToSocial, which is Mastodon-compatible, and it’s pretty sleepy over there. There’s a little anti-AI sentiment in the Fediverse too, and content discovery is hard. Finding posts from the kind of people I actually want to interact with is a struggle. Even LinkedIn, where you’d expect more engagement from professionals working in AI, is only a little better. All of it left me feeling this week like I might have been better off as a hermit. I don’t know what to do with that yet. But it’s there.

Next up: a model router. Every time new models come out and I switch over to them, moving all of my different agent harnesses onto the new models is a lot of work. The plan is to put something like LiteLLM in my home lab as a routing layer between all of my harnesses and the models they use. Set the new models in that one place, and every harness picks them up transparently. They won’t even know something changed. They’ll just start benefiting from the new models right away.

August 22, 2026#

Three releases in one day. The agent-skills v0.7.0 release is the big one: thirty-one new skills, upgrades to more than twenty existing ones, and a repo restructure that flattens everything to the root so it’s actually pleasant to clone and browse. I wrote the whole thing up in agent-skills v0.7.0: Thirty-One New Skills and One Flat Repo. But it wasn’t alone: SlopSearX v0.3.0 shipped an MCP server for AI agents, durable research jobs, explainable cost- and coverage-aware routing, explicit media search, and job-board engine adapters for Greenhouse, Ashby, and Lever. And GroktoCrawl v0.13.0 added autonomous CAPTCHA detection and recovery, a deterministic SlopSearX contract twin for testing, and adopted firecrawl-anydoc as its document conversion engine.

Four Groktopus articles this week. AI Governance Is Human Work makes the case that an AI-governance skill can advise and challenge, but accountable humans still own the decisions. AI Spending Caps Become Permission Systems looks at how blunt per-employee AI wallets suppress exactly the experimentation they’re supposed to fund. You Want a Software Factory? Ready Your Repositories First. leans on DORA’s finding that 90% of developers use AI but 30% report little or no trust in what it generates, then builds the repository controls that close that gap. And Digital Twin Engineering: Test APIs Without the Bill is a plain-English walkthrough of modeling API workloads before you spend real money on them.

Open Source Projects#

I’m maintaining a few things right now:

• hermes-cashew. A thought-graph memory provider for Hermes Agent. Gives AI agents persistent, structured memory across sessions instead of starting from zero every conversation. Apache 2.0, built in Python, getting real use.

• hermes-SkillOpt. Controlled skill optimization for Hermes Agent, derived from Microsoft Research’s SkillOpt.

• groktocrawl. Self-hosted web scraping and research stack. What started as a Firecrawl alternative turned into something more capable: multi-page crawling, JS-rendered content, LLM-friendly output. Runs on Docker, works with any agent framework.

• SlopSearX. Cloud-native, stateless, AI-agent-first meta search engine. Drop-in SearXNG replacement for the GroktoCrawl stack.

• artifact-pyramids. Progressive disclosure for what AI agents produce: Summary, Analysis Collection, and Dossiers. The Artifact Pyramid is a structured methodology for organizing AI agent research outputs across three layers of increasing depth. Just as progressive disclosure governs how we feed agents context, the Artifact Pyramid governs what they produce, enabling downstream agents and humans to consume at the depth they need.

• hugo-cyberdelic. A cyberdelic Hugo theme: 1969 psychedelic concert poster meets cyberpunk holographic future.

• agent-skills. A growing library of reusable skill modules for Hermes Agent. Reference patterns, workflow templates, and standalone skills I’ve extracted from real use. If you’re running Hermes, there’s probably something useful in here.

• hermes-profiles. Curated Hermes Agent profiles for specialist swarms. Opinionated, Hermes-optimized role definitions for delegation, kanban pipelines, and multi-agent orchestration.

• programming-principles. Programming principles distilled from 14 classic software engineering books, packaged as an Agent Skill. The lessons without the 400-page commitment.

• hermes-claude-import. Import Claude.ai conversation exports into Hermes Agent memory providers. Community collaboration, open source.

• hermes-council. A multi-agent debate system for Hermes Agent, now running as a native plugin with typed contracts and structured inference. Composes panels of expert personas to surface hidden assumptions and stress-test ideas. MIT licensed, open source from day one. I use it for everything from editorial review to strategic analysis.

What I’m Reading (or Rather, Listening To)#

(stay tuned)

This is my now page, inspired by Derek Sivers. It’s a snapshot of what I’m focused on at this point in my life. I promise not to let another four months go by before the next update.

More research on my about Now Pages note.