◆ WRITING

SIGNAL LOG

Field notes on AI systems, agent interfaces, and the products emerging between them.

06 articles

01
META MUSEAI AGENTSCREATOR OS

Why I’m Testing Meta Muse: Where the Access Advantage Might Be

My early thoughts on testing Meta Muse for research, useful context, and the workflows I’m building with Creator OS, Content Radar and TeamBot.

Oct 19, 202610 MIN READ
02
WORLD MODELSSPATIAL AIAI AGENTSFDE

From AI Agents to World Models

The future of AI won't only think. It will simulate and act. Exploring world models, spatial AI, and the next step beyond agents.

Oct 15, 202614 MIN READ
03
CLAUDE CODEMODSAGENT UXTEAMBOT

Claude Code stopped being a tool and became a workspace you can program.

Exploring Claude Code Mods, programmable agent interfaces, and my Octo Quest and TeamBot Mission Control experiments.

Oct 5, 202619 MIN READ
04
JEVTYPESAFE AIAGENT ROUTINGCREATOR OS

Not every AI problem needs the smartest AI.

Jev doesn't write, code, or brainstorm. It only makes decisions — and thinking about why that's useful changed how I'm planning Creator OS's architecture.

Sep 19, 202616 MIN READ
05
CLAUDE CODECODEXCONTEXT ENGINEERINGAI AGENTS

Your usage isn't disappearing. You're paying the agent to do unnecessary work.

Twenty techniques for cutting Claude Code and Codex costs without giving up quality — and why most of them make the output better.

Sep 9, 202618 MIN READ
06
WEBMCPAI AGENTSPRODUCT DESIGN

WebMCP and the Agent Economy: the visitor isn't human anymore.

The paradigm shift from attention to resource-usage rights—and what agent-ready websites mean for the products we build.

Sep 3, 202621 MIN READ