overnight-dev
Run Claude autonomously for 6-8 hours overnight using Git hooks that enforce TDD - wake up to fully tested features
View on GitHubTable of content
Run Claude autonomously for 6-8 hours overnight using Git hooks that enforce TDD - wake up to fully tested features
Installation
npx claude-plugins install @jeremylongshore/claude-code-plugins-plus/overnight-dev
Contents
Folders: agents, commands, examples, scripts, skills
Files: LICENSE, PLUGIN_SUMMARY.md, README.md
Documentation
Go to bed. Wake up to fully tested features.
Run Claude autonomously for 6-8 hours overnight using Git hooks that enforce test-driven development. No more “should work” - only “does work.”
The Problem
You write code all day. Tests fail. You debug. Repeat. Progress is slow.
What if Claude could keep working while you sleep?
The Solution
Overnight Development turns Claude into an autonomous developer using Git hooks that enforce TDD:
- Git hooks block commits until all tests pass
- Claude can’t commit broken code - hooks enforce quality
- Automatic iteration - debug, fix, retry until green
- You wake up to fully tested, production-ready features
Quick Start (2 minutes)
# 1. Install the plugin
/plugin install overnight-dev@claude-code-plugins-plus
# 2. Setup in your project
/overnight-setup
# 3. Start coding
# Every commit now requires passing tests!
That’s it. Hooks are installed. Quality is enforced. Let Claude work overnight.
How It Works
Traditional Development
Write code → Hope it works → Find bugs later → Repeat
Slow, error-prone, stressful
Overnight Development
Write test → Write code → Hooks enforce tests → Commit only when green
Fast, reliable, autonomous
The Magic: Git Hooks
Pre-commit hook:
# Runs before EVERY commit
1. Lint your code
2. Run your tests
3. Check coverage
4. If anything fails → Commit blocked
5. Claude sees the error → Fixes it → Tries again
Commit-msg hook:
# Ensures quality commit messages
- Enforces conventional commits format
- feat: fix: docs: test: etc.
- Makes git history readable
The Result
Claude can’t commit until all tests pass. So it keeps working until they do. Overnight. While you sleep.
Real-World Example
9 PM: Start overnight session
/overnight-setup
# Task: "Build JWT authentication with 90% test coverage"
What Claude does overnight:
- 10:15 PM - Write failing auth tests (TDD)
- 10:45 PM - Implement JWT signin (tests still failing)
- 11:30 PM - Debug token generation (commit blocked, keeps trying)
- 12:15 AM - Tests passing! Commit succeeds
- 1:00 AM - Add middleware (write tests first)
- 2:30 AM - Integration tests (debugging edge cases)
- 4:00 AM - All tests green (Coverage: 94%)
- 5:30 AM - Add docs, refactor, still green
- 7:00 AM - Session complete
7 AM: You wake up to:
- 47 passing tests (0 failing)
- 94% test coverage
- Clean conventional commit history
- Fully documented JWT authentication
- Production-ready code
Time saved: 6-8 hours of your life
Features
Test-Driven Development Enforcement
Git hooks force TDD:
- Write tests first
- Implementation after
- Commit only when green
- No shortcuts, no broken code
Autonomous Debugging
When tests fail:
- Claude reads the error
- Analyzes the problem
- Forms a hypothesis
- Makes a fix
- Tries to commit again
- Repeat until green
You never have to intervene. Just check progress in the morning.
Progress Tracking
Watch overnight sessions in real-time:
# View the log
cat .overnight-dev-log.txt
[22:15] Session started: JWT Authentication
[22:20] Tests: 12/12 passing
[23:10] Tests: 18/18 passing - Auth routes done
[00:30] Tests: 24/24 passing - Middleware complete
[02:15] Tests: 35/35 passing - Integration tests done
[04:00] Tests: 47/47 passing - Coverage 94%
[06:45] SESSION COMPLETE
️ Flexible Configuration
Works with any test framework:
Node.js:
{
"testCommand": "npm test",
"lintCommand": "npm run lint"
}
Python:
{
"testCommand": "pytest --cov=.",
"lintCommand": "flake8 ."
}
Rust, Go, PHP, Ruby - All supported!
Smart Agent Guidance
Includes overnight-dev-coach agent:
- Guides you through setup
- Plans overnight tasks
- Debugs failing tests
- Tracks progress
- Celebrates success
Activate by mentioning “overnight development” or asking about autonomous coding.
Installation & Setup
Prerequisites
Before installing, you need:
Git repository - git init in your project
Test framework - Jest, pytest, cargo test, etc.
At least 1 passing test - Hooks need something to run
Linter configured - ESLint, flake8, clippy, etc.
Install the Plugin
# Add the Claude Code Plugin marketplace
/plugin marketplace add jeremylongshore/claude-code-plugins
# Install overnight-dev
/plugin install overnight-dev@claude-code-plugins-plus
Setup in Your Project
# Run the setup command
/overnight-setup
This creates:
.git/hooks/pre-commit- Tests and linting.git/hooks/commit-msg- Conventional commits.overnight-dev.json- Configuration
Verify It Works
# Try a commit (should run tests)
git commit --allow-empty -m "test: verify hooks"
You should see:
Overnight Dev: Running pre-commit checks...
Run
...(truncated)
## Included Skills
This plugin includes 1 skill definition:
### overnight-development
> Automates software development overnight using git hooks to enforce test-driven Use when appropriate context detected. Trigger with relevant phrases based on skill purpose.
<details>
<summary>View skill definition</summary>
# Overnight Development
## Overview
This skill automates software development overnight by leveraging Git hooks to enforce test-driven development (TDD). It ensures that all code changes are fully tested and meet specified quality standards before being committed. This approach allows Claude to work autonomously, building new features, refactoring existing code, or fixing bugs while adhering to a rigorous TDD process.
## Prerequisites
- Access to project files in {baseDir}/
- Required tools and dependencies installed
- Understanding of skill functionality
- Permissions for file operations
## Instructions
1. Identify skill activation trigger and context
2. Gather required inputs and parameters
3. Execute skill workflow systematically
4. Validate outputs meet requirements
5. Handle errors and edge cases appropriately
6. Provide clear results and next steps
## Output
- Primary deliverables based on skill purpose
- Status indicators and success metrics
- Generated files or configurations
- Reports and summaries as applicable
- Recommendations for follow-up actions
## Error Handling
See `{baseDir}/references/errors.md` for comprehensive error handling.
## Examples
See `{baseDir}/references/examples.md` for detailed examples.
## Resources
- Official documentation for related tools
- Best practices guides
- Example use cases and templates
- Community forums and support channels
</details>
## Source
[View on GitHub](https://github.com/jeremylongshore/claude-code-plugins-plus-skills)