Claude Code Gains Official Built-in Plugin 'You should Know' That Flags What You Might Miss via a Side Agent

Anthropic's Claude Code team announced a new official built-in plugin, 'You should Know (cc-plugin-you-should-know@builtin)', on October 3, 2026. Built on the C

tau · October 3, 2026

#ClaudeCode #Anthropic #AICoding #DeveloperTools #ClaudePlugins

Claude Code Gains Official Built-in Plugin 'You should Know' That Flags What You Might Miss via a Side Agent

Anthropic's Claude Code team (@ClaudeDevs) announced a new official built-in plugin for Claude Code, 'You should Know', on October 3, 2026. It scans Claude's output for important information developers might miss to help keep them in the loop, and it activates with a single command: /plugin enable cc-plugin-you-should-know@builtin.

The plugin ships as a built-in extension that runs inside a Claude Code session. There is no third-party install path involved — developers turn it on directly from a session, which is the most practical point of this release.

A Mod-Built Side Agent That Watches the Main Output

'You should Know' is built on the 'Claude Mods' architecture. Mods are the extension type that lets plugins modify deeper agent behavior in Claude Code.

  • Dedicated side agent: it spins off a side agent whose only job is to observe Claude's output, flagging things the developer or Claude might miss.
  • Enable command: type /plugin enable cc-plugin-you-should-know@builtin in a Claude Code session.
  • Version floor: supported on Claude Code v2.1.287 and later.
  • Off by default: it ships disabled, so users must turn it on explicitly.

The team pointed readers to its official blog post on Mods (Getting Started with Claude Code Mods, https://claude.dev/blog/getting-started-with-claude-code-mods/) for more background.

Who It Helps and What to Check Before Enabling

It suits developers running long sessions where skimming every line of Claude's verbose output is impractical. A second reader pass over the same output targets exactly that failure mode: important cautions or key details slipping by unnoticed.

One caveat follows directly from the design: running a background side agent means additional model calls and token usage. The announcement does not specify what the side agent treats as important, how it surfaces findings, or the exact extra cost. Mods run with the user's permissions, so even for an official built-in, confirming its state in the plugin list before enabling is the safe move.

Sources