Library
AI Agent Night Shift Playbook
Kit files are in English.
Set up overnight agent work with budget caps, stop conditions and a morning report.
Yearly: every kit included. Monthly: one new kit per billing cycle.
View full-size sample ↗Access, downloads and refund policy
Yearly: every available kit and bundle, unlimited downloads.
Monthly: one new standalone kit per billing cycle, no bundles. Re-downloads are free and do not use a new claim.
Monthly subscriptions started before February 28, 2026 keep all-access, including bundles and unlimited downloads.
Monthly access currently allows three downloads per kit. Need another? Contact joozio@hey.com.
Download access while subscribed. Keep the files you download.
Already bought this? Your download link still works.
How it works
Set up an AI agent that works while you sleep. Real configs, Git preflight, bounded autonomy rules, memory systems, resource safety, and cost optimization from production use.
100+ automated sessions completed in production. Now updated with bounded night shift rules.
Latest update: 2026-05-11
Added bounded night shift rules, Git clean-state preflight, and resource safety checks.
What you get
Package includes
- •Practitioner guide (8,000+ words, not theory)
- •Wake script templates (bash, tested in production)
- •bounded-night-shift.md: May 2026 safety model for unattended sessions
- •resource-safety-checklist.md: CPU, memory, thermal, and local-model stop conditions
- •LaunchD plist + cron + systemd configs
- •CLAUDE.md template for autonomous agents
- •Memory system templates (3-tier)
- •Morning report script template
- •Error registry template
- •Cost calculator reference
- •3 complete night shift configs (research, deployment, content)
FAQ
What do I need to run this?
Claude Code, an Anthropic API key, and a computer that stays on overnight (Mac, Linux, or a cloud VM). The guide covers setup for all three.
How much do night shifts cost?
Typically $2-5 per night using optimized model selection. The guide shows how to pick the cheapest model that works for each task type.
Is this the actual system from the blog?
Yes. Same architecture, adapted as reusable templates. The blog post describes what it does. This playbook shows exactly how to build your own.
Related reading on Digital Thoughts:
- How to Use Git(hub) When You Are Building with AI (Basics)
- The Bounded AI Agent
- How to (Almost) Fry Your AI Agent (and Your Mac Mini)
- How to Build Your First AI Agent (Basics)
- I Gave My AI Agent Its Own Computer. Here's Every Lesson From 72 Hours of Migration
- Building an AI Agent That Runs Night Shifts
- I Submitted to a Mistral Hackathon at 10pm. My Agent Built the App While I Slept
- When AI Meets Reality: What Actually Happened When I Let My Agent Run Wild