Sign in Subscribe

Topic

Claude

A collection of 18 issues

Claude Duet gets your things done. The harder question is who gets our things done.

Anthropic's Claude Duet is the proactive assistant many of us were waiting for. For a solo operator it could be the back office. For a team of five, it raises a harder question: when does a personal assistant become an orchestrator?

Credentials Without Chaos: How to Use Claude Safely With Sensitive Data

Claude needs access to your tools, not to your passwords. Here's how I set it up: Cloudflare Workers in the cloud, with a password vault as the on-computer fallback.

Claude Code mods vs skills: a skill is a suggestion, a mod is a guarantee — and what I would hard-block first

Mods run inside the Claude Code harness and can refuse a tool call outright, where a skill can only advise. The scoping rule that follows, the four I would actually build, and the security posture they demand.

File timestamps can't see staleness that comes from outside the folder: what an agentic housekeeping routine needs that mtime doesn't give it

My weekly routine rebuilds stale project summaries by comparing file modification times. It ran clean — and missed a summary made wrong by a decision taken in a different folder. Local change detection is not correctness.

Microsoft's new Copilot looks just like Claude. That's not the part that matters.

Microsoft relaunched Copilot with Home, Code and Autopilot. It looks a lot like Claude, but the real story is distribution: Microsoft is already inside the business.

I tested my new Claude skill against an agent that didn't have it — and the control group found the saving my skill missed

The skill was 35% faster and used 12% fewer tokens. The baseline agent, with no instructions at all, found €10.51 the skill walked straight past. Why the no-skill run is a second opinion, not just a scoreboard.

Meta's Muse, one week in: the model isn't the best, and the integrations make up for it

A week of real Muse users on Threads: everyday errands, not code, and one line that sums it up. The model isn't the best, the integrations make up for it. Why that means your context has to be tool-agnostic.

Claude just merged Chat and Cowork — the decision that disappeared, and the caveat that didn't

Chat and Cowork are now one conversation with a full agent runtime underneath. What changes for operators: no more routing, deliverables inside the chat — and the local-tasks caveat most coverage skips.

My assistant created the same task three days running. The fix has a name: idempotency.

A tiny automation bug taught the same lesson payment systems learned years ago: the question isn't whether it runs, it's what happens when it runs again.

A context window is working memory, not a hard drive. Feel that, and you manage AI differently.

The AI doesn't remember everything you've told it. It has working memory, and it's finite — here's the cheap trick I use to hear it filling up, and the fix when it does.

Work with me →×