Clockwork
Claim this toolSchedule and run AI coding agents safely in sandboxed git worktrees
Visit WebsiteFreemiumVisit Website
Tracked since2026
The Bottom Line
Entry price
Free plan available, paid tiers above
Biggest pro
Strong sandbox isolation protects your code and credentials during agent runs
Biggest con
macOS only, and jobs run only while the Mac is awake
TL;DR - Clockwork
- Schedules coding agents via a calendar, cron, or webhooks.
- Runs each agent in a sandboxed git worktree with strict credential protection and enforced budget/time caps.
- Files a structured, searchable report after every run, including summary, diffstat, cost, approvals, and transcript.
Pricing: Free plan available
Best for: Growing teams
What is Clockwork?
Clockwork is a macOS calendar for your coding agents. It lets you schedule AI coding agents like advanced models as calendar events, one-off, recurring, or webhook-triggered. Each run executes inside a sandboxed git worktree with a macOS Seatbelt profile, so your main checkout, SSH keys, cloud credentials, and shell history stay protected. You set dollar, turn, and time caps, and every run files a structured report into your inbox with a summary, branch diffstat, cost, approvals, and a full searchable transcript.
Designed for developers who want automated code chores to happen reliably without babysitting, Clockwork includes thirteen pre-built agent specialists, from bug hunting and dependency surgery to code review and release readiness, or you can define your own profiles. Scheduling respects real calendar recurrence (RRULE or cron) with daylight-saving handling and missed-run policies. It's local-first: no cloud account, no telemetry, and your provider keys live in the macOS Keychain.
Pros & Cons
Pros
- Strong sandbox isolation protects your code and credentials during agent runs
- Detailed, searchable reports after every run make review easy
- No cloud dependency, everything runs locally with keys in the macOS Keychain
Cons
- macOS only, and jobs run only while the Mac is awake
- Reproducible builds are not yet supported, so downloaded builds can't be independently proven to match source
Key Features
Calendar-based scheduling with RRULE or cron, daylight-saving aware and with missed-run policiesSandboxed execution using git worktrees and macOS Seatbelt profilesCredential deny-list protecting .ssh, .aws, .gnupg, and shell historySupports Claude Code, Codex CLI, OpenCode, Hermes Agent, and OpenAI-compatible providersDollar cap, turn cap, and wall-clock timeout enforced by a supervisorAuto-generated reports with summary, branch diffstat, cost, approvals, and full transcriptThirteen built-in agent specialists plus custom profile creationWebhook and GitHub triggers with HMAC verification
Pricing
Freemium
Clockwork offers a generous free tier with optional paid upgrades for advanced features.
Reviews
$99Free with your review
Write a reviewReview Clockwork, get a free AI guide
Share your experience and we will send you Improve Your Thinking Patterns Using ChatGPT, free.
Best Clockwork Alternatives
Top alternatives based on features, pricing, and user needs.
AutoGPTFreemium
Autonomous AI agent that breaks down and completes complex goals
Composio MCPFreemium
Empower AI agents to act across 1000+ apps securely
Open AgentsFreemium
Unified workspace for AI agents to collaborate, share context, and execute tasks.
CrewAIFreemium
Multi-agent orchestration framework for autonomous AI teams
AGNT.HubPaid
Build and manage secure, private AI agents with custom skills
BabyAGIFree
An experimental framework for building self-building autonomous agents with a function management system.
Still deciding?
Most buyers shortlist 2 or 3 tools before committing. Pull a side-by-side comparison or browse the full alternatives shortlist below.
Explore More
Clockwork FAQ
How does Clockwork help with automating code review tasks?
Clockwork lets you schedule a pre-built code review specialist as a calendar event, running in a sandboxed git worktree. After each run, it files a structured report with a summary and diffstat, making review easy without manual overhead.
How does Clockwork differ from CrewAI?
Unlike CrewAI's multi-agent orchestration, Clockwork focuses on sandboxed execution inside macOS git worktrees with a Seatbelt profile, protecting your main code and credentials. Clockwork runs locally with no cloud dependency and keys stored in macOS Keychain, while CrewAI often relies on cloud-based coordination.
What are the main trade-offs when using Clockwork?
Clockwork runs only on macOS and requires the Mac to remain awake for scheduled agent jobs. Additionally, reproducible builds are not yet supported, so downloaded builds cannot be independently proven to match the source code.
What kind of user benefits most from Clockwork?
Developers who want automated code chores to happen reliably without babysitting benefit most from Clockwork. It suits individual developers or small teams who prefer a local-first tool with strong sandbox isolation and detailed reports.
Does Clockwork include a free tier?
Yes, Clockwork is available on a free tier, with paid plans for more usage and features. This allows developers to try the scheduling and sandboxed agent runs before committing to a paid plan.
How does Clockwork ensure my code and credentials stay safe during agent runs?
Each agent run executes inside a sandboxed git worktree with a macOS Seatbelt profile, isolating it from your main checkout, SSH keys, cloud credentials, and shell history. Your provider keys are stored in the macOS Keychain, and no cloud account or telemetry is used.
Can I create my own agent profiles in Clockwork?
Yes, Clockwork includes thirteen pre-built agent specialists, and you can also define your own custom profiles. This flexibility allows you to tailor automated chores like bug hunting, dependency surgery, or release readiness to your specific workflow.
How does Clockwork handle recurring schedules and daylight saving time?
Scheduling respects real calendar recurrence using RRULE or cron, with proper daylight saving handling and policies for missed runs. This ensures code chores run reliably at the intended times without manual intervention.
Source: vimoxshah.github.io