nightshift

Automate an autonomous AFK development loop processing specs and bugs.

2|Updated Feb 23, 2026
One-click install
npx skills add https://github.com/jonmumm/skills --skill nightshift-jonmumm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nightshift
Source: https://github.com/jonmumm/skills/tree/main/nightshift
Command: npx skills add https://github.com/jonmumm/skills --skill nightshift-jonmumm

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires claude, codex, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Nightshift automates an autonomous AFK development loop, handling specs and bugs end-to-end so you can ship features while away or on breaks.

Core Features & Use Cases

  • Automated backlog discovery and prioritization (Bugs first, then specs)
  • Acceptance-test-first workflow across Web, iOS, and React Native
  • Configurable duration and per-run prompts for AI agents
  • Morning briefing, NOTICED.md, and lesson persistence for cross-run learning

Quick Start

Run the nightshift script with your project path to start an autonomous AFK development loop.

Frequently Asked Questions about nightshift

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate an AFK development loop to process specs and bugs end-to-end?

You can automate an AFK development loop by running a script that reads project docs and structure to prioritize bugs and specs, executing tasks autonomously while you are away. It drives development end-to-end across web, iOS, and React Native projects.

Does autonomous acceptance testing work with React Native and iOS projects?

Yes, autonomous acceptance testing works with React Native and iOS projects. The workflow applies an acceptance-test-first approach across web, iOS, and React Native, prioritizing tests to validate features developed during the automated loop.

How do I configure duration limits and AI agents for an automated workflow?

You can configure duration limits and AI agents by providing a duration hint and per-run prompts when starting the autonomous loop. The workflow supports optional Claude or Codex agents to drive the development tasks within the specified timeframe.

Can I use CLAUDE.md knowledge bases to drive automated end-to-end development tasks?

Yes, you can use CLAUDE.md knowledge bases to drive automated development tasks. The workflow reads CLAUDE.md files alongside project docs and structure to inform and guide the AI agents processing your backlog.

What is the best way to review autonomous development results after an AFK run?

The best way to review autonomous development results is through the morning briefing and NOTICED.md file generated after the run. It also persists lessons learned for cross-run learning, ensuring continuous improvement across future automated loops.