loop

Automate iterative evaluation-fix-validate cycles for code quality improvements.

2|Updated Feb 16, 2026
One-click install
npx skills add https://github.com/donygeorge/claude-toolkit --skill loop-donygeorge
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: loop
Source: https://github.com/donygeorge/claude-toolkit/tree/main/skills/loop
Command: npx skills add https://github.com/donygeorge/claude-toolkit --skill loop-donygeorge

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Iterative evaluate-fix-validate loop helps teams converge on code quality by repeatedly assessing findings, applying fixes, and validating results until stability.

Core Features & Use Cases

  • Iterative evaluation: fresh agents assess correctness, style, safety, and test coverage across the scope.
  • Convergence-driven fixes: fixes are selected by severity, then validated and committed per iteration.
  • Controlled scope evolution: scope grows within per-iteration and total file limits to prevent explosion.

Quick Start

Run /loop with your feature scope to start iterative improvement on your codebase.

Frequently Asked Questions about loop

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

FAQPage Schema
How do I automate iterative code quality improvements across multiple files?

An iterative evaluation-fix-validate loop works by repeatedly assessing codebase findings, applying fixes selected by severity, and validating results per iteration. This cycle continues until the code quality converges on a stable state across the defined feature scope.

How do I start an iterative code evaluation and fixing loop for my project?

Start iterative code evaluation by running the loop command with your defined feature scope. This initiates the automated cycle of assessing findings, applying severity-selected fixes, and validating results to systematically improve your codebase quality.

Do I need a specific configuration file to run automated code quality loops?

Yes, automated code quality loops require a configured toolkit.toml file and an accessible codebase. These prerequisites allow the system to track scope, iterations, findings, and state across multiple runs to ensure controlled convergence.

How does iterative code validation handle scope growth and complex findings?

Iterative code validation controls scope growth using per-iteration and total file limits to prevent explosion. Complex findings are deferred or branched, allowing the primary loop to converge on code quality without being blocked by edge cases.

What is the best way to converge on code quality for a large feature scope?

The best way to converge on code quality for a large feature scope is using an automated evaluate-fix-validate loop. It selects fixes by severity, controls scope evolution with file limits, and commits validated improvements per iteration until stability.