Docs

Runtime

See every cron, hook, and background job - when it runs, what it touches. The non-request side of the system: scheduled jobs, lifecycle hooks, queue workers, and what each one writes.

See it in motion

Where to find it

  • Desktop: localhost:4000, then Memory in the sidebar, then Reference index under All tools, in the Lessons, wiki and patterns group, which lists it.
  • Hosted: desktop only.
  • Keyboard: ⌘ K, then type “Runtime”.
  • On disk: .claude/brain/runtime.md

What it does for you

Every cron + hook in one inventory.When you wonder “what touched this file overnight,” the answer is here. Each row: schedule, entry point, what it writes, where its log goes.
Hooks are first-class, not buried in settings.Claude Code hooks (SessionStart, PreToolUse, etc.) appear with what they enforce and where their script lives.
Background jobs don't get to be invisible.Telemetry pullers, brain-reflection runs, sync jobs - each one's schedule and side-effects are written down. Surprise jobs cost more than they save.

Configure

Crons are usually GitHub Actions or platform-scheduled functions. Hooks live in .claude/settings.json. Add a job → add the row here in the same PR.

Use it well

Before adding a new cron, scan the list - there's often an existing job that can do the work. When adding one, log the schedule, the side-effects, and the failure behavior. Quarterly: kill the jobs that aren't earning their slot.

Read more

Last updated