broken-window-check

Run end-to-end tests on recently completed features at session start.

746|130|Updated Jun 30, 2026

Search Tags:#code quality #bug prevention #end-to-end testing #feature validation #shift workflow

Ready to run

npx skills add https://github.com/Archive228/loopkit --skill broken-window-check

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: broken-window-check
Source: https://github.com/Archive228/loopkit/tree/main/skills/broken-window-check
Command: npx skills add https://github.com/Archive228/loopkit --skill broken-window-check

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill ensures that features marked as done are truly operational before new work begins, preventing the accumulation of technical debt.

Core Features & Use Cases

Smoke-Test Recent Features: Before starting new work, the skill runs end-to-end tests on the last "completed" feature.
Detect and Address Issues: If a feature fails, the skill prompts the user to revert changes, re-open the feature, and fix the issue.
Use Case: It is designed for shift-driven workflows where work can be left in an incomplete state from one session to another.

Quick Start

Execute the 'broken-window-check' command after reading your progress notes to validate the most recent completed feature.

Frequently Asked Questions about broken-window-check

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

FAQPage Schema

How do I verify that a recently completed feature actually works before starting new development?▼

To verify a completed feature works before new development, you run end-to-end test scripts to perform runtime verification of its functionality. This validates that the feature is operational and prevents building new code on top of broken functionality.

Why should I run an end-to-end test at the start of a development session?▼

Running an end-to-end test at the start of a development session ensures continuous integration practices by catching incomplete or broken features from previous shifts. This prevents the accumulation of technical debt before you write any new code.

What should I do if end-to-end testing fails for a feature marked as done?▼

If end-to-end testing fails for a feature marked as done, you should revert the recent changes, re-open the feature ticket, and fix the issue immediately. This ensures you do not accumulate technical debt by building on broken code.

How do I validate feature functionality in shift-driven workflows where work is left incomplete?▼

To validate feature functionality in shift-driven workflows, execute a broken-window check after reading your progress notes to run end-to-end tests on the most recent completed feature. This confirms the previous session's work is truly operational before starting new tasks.

Can I perform feature validation without writing custom test scripts?▼

No, this feature validation requires the execution of existing test scripts for runtime verification. You must have end-to-end tests in place to automatically detect and address issues with recently completed functionality.

What is the best way to prevent building on broken code in continuous integration?▼

The best way to prevent building on broken code in continuous integration is to perform automated end-to-end testing of recent features at the start of every development session. This catches failures early and prompts you to revert and fix issues before proceeding.