This is the Claude Code team catching up to something OpenClaw has had for a while — and shipping it as a native, first-class feature.
What It Does
You define a task and a schedule. Claude Code runs the task at the specified intervals, on your local machine, using whatever tools and context you've set up.
Examples:
- Daily morning briefing: summarize overnight news, check your project's GitHub issues, draft your day's priorities
- Hourly code health check: scan recent commits for patterns, flag anything that looks concerning
- Weekly report: pull data from your tools, compile a summary, format it for your review
The tasks run locally — on your machine, with your credentials, with access to your local files — not on Anthropic's servers.
The OpenClaw Comparison
OpenClaw's heartbeat and cron system has enabled this kind of scheduled autonomous work since early versions. The difference: OpenClaw's system is more flexible (full shell access, external API calls, multi-channel notifications) and Claude Code's is more integrated (native to the code-focused workflow, simpler to set up for developers).
For developers who live in Claude Code but want scheduled automation, the native integration is cleaner than setting up a separate OpenClaw instance.
What This Enables
The scheduled tasks feature moves Claude Code from a tool you invoke to infrastructure you configure once and rely on. Your agent doesn't wait for you to show up — it maintains your codebase, runs your reports, monitors your systems, whether you're at your keyboard or not.
For solo founders running engineering operations alone, this is meaningful infrastructure.