Clockwork
Schedule AI agent tasks on a real calendar — book, run, review, and repeat with isolated sandbox execution
| What is it | Schedule AI agent tasks on a real calendar — book, run, review, and repeat with isolated sandbox execution |
|---|---|
| Pricing | Free |
| Free tier | Yes |
| Platform | Desktop Application |
| Best for | Scheduling recurring AI code maintenance tasks like dependency updates and test triage, Running unattended agent jobs with safety oversight and audit trails |
| Domain registered | 2013 |
Data updated Sept. 19, 2026
What does Clockwork do?
Clockwork is a macOS desktop app that brings AI agent work onto a real calendar. Instead of managing cron jobs, shell scripts, or CI pipelines for recurring agent tasks, you book work on a month-grid calendar. Clockwork runs those jobs in isolated sandboxes at the scheduled time and delivers a readable report with a summary, diffstat, cost, and transcript.
The app works with the AI coding engines you already use: Claude Code, Codex CLI, OpenCode, or Hermes Agent. You pick the provider per task without any setup changes. Each run gets its own macOS Seatbelt sandbox and git worktree, so file writes are locked to scope and credentials stay unreadable. Human-in-the-loop gates pause risky actions and fail safe if you don't respond. Budget caps, turn limits, and wall-clock timeouts are enforced by a supervisor, not the model's promises. The full security boundary — seatbelt profiles, deny lists, and allow lists — is published under Apache-2.0 along with tests that prove the sandbox actually blocks secret access.
Clockwork is free during beta and brings your own API keys — there's no markup on provider charges. It's built by terminal people for terminal people, with a ⌘K command palette, global shortcuts, and instant inbox search. Developers who need to schedule recurring code maintenance tasks like dependency updates, test triage, or repo health monitoring will find ready-to-run specialist agents. Engineering teams that want audit trails and policy enforcement across agent usage will want the coming Pro tier. The honest limitation is that jobs run only while your machine is awake, so overnight tasks need an always-on Mac.
Key features
What makes it stand outWho is Clockwork for?
Who benefits most from this toolTrust & presence
Alternatives in Developer Tools
Cloud scheduler for AI agents — schedule LangChain, CrewAI, and webhook tasks with one CLI command
Native macOS app to manage Codex and Claude Code CLI sessions in one workspace with split panes, worktrees, and embedded simulators.
Run Claude Code, Codex, Cursor and other AI coding agents in one desktop, terminal, and iOS app
IT control plane for internal AI agents — approve, monitor, and audit agents built in any tool from one dashboard
Terminal UI to spawn, monitor, and review multiple AI coding agent sessions from one keyboard
Desktop app that monitors your coding agents (Cursor, Claude Code, Codex) and suggests improvements to their rules and skills.
Native macOS app that schedules and automates Claude Code tasks to run in the background at specified times.
Cloud VM for coding agents — run Codex and Claude Code threads in parallel, each on its own Linux box
Similar tools
Desktop app for using 50+ LLMs with your own API keys, sharing setups with your team.
AI-powered calendar assistant that automatically optimizes your schedule to create focus time and reduce meeting conflicts.
AI-powered time tracker that automatically logs your work hours from computer activity