2026-10-04 You Should Know - a second pair of eyes on long Claude Code runs

Canonical version: 2026-10-04 You Should Know - a second pair of eyes on long Claude Code runs.

On 2 October, @ClaudeDevs announced You Should Know, a new built-in plugin for Claude Code. The post passed 1.5 million views. It's a small feature, and it's the first built-in I'd turn on.

What it does

On longer tasks, a side agent scans Claude's output for important information you might miss and shows a note above the prompt when it finds some. Anthropic says it's there "to help keep you in the loop".

It ships as a built-in mod, disabled by default. Turn it on with:

/plugin enable cc-plugin-you-should-know@builtin

Why it matters

On a long autonomous run, Claude can print hundreds of lines, and you won't read all of them. If one of those lines mattered (say, a skipped test), you'll likely only see the summary at the end.

My take on the Opus 5.5 prompting tips came from the same concern: give it the whole task, keep your hand on the brake. You Should Know doesn't brake for you, and it doesn't replace permission prompts or explicit stop lines. It makes it more likely that you notice WHEN to brake.

It's also a filter. If the side agent decides something isn't important, you get no note, so keep reviewing the diff.

And the side agent uses tokens. Anthropic doesn't say how many, so check your usage after a few long sessions.

How it fits with the other built-in mods (and how mods work in general) is in my Claude Code Mods note.

That's it for today! ✨

References


About Sébastien

Ready to get to the next level?

Found this valuable? Share it with someone who needs it.

Join 6,000+ readers. Get practical systems for knowledge & AI. Free.

Subscribe ✨

Free: Knowledge System Checklist

A clear roadmap to building your own knowledge system. Subscribe and get it straight to your inbox.

6,000+ readers. No spam. Unsubscribe anytime.

Subscribe