How does the skill handle discovered errors?

The skill categorizes issues by severity: Critical (functional impact), Major (consistency issues), and Minor (small improvements), requiring all Critical issues to be resolved before committing.

What files does the Iterative Review skill work best with?

It is specifically optimized for documentation and configuration files, such as Markdown, JSON, and YAML, found in .opencode, .claude, and .github directories.

What are the four primary review criteria used?

The review focuses on Consistency (formatting and naming), Completeness (missing data), Correctness (logic and paths), and Integration (cross-references between files).

What happens if a problem cannot be fixed automatically?

If the skill reaches a limit of 5 iterations without resolving the issues, it is designed to escalate the problem to the user to re-evaluate the implementation strategy.

OpenCode Iterative Review

Name: OpenCode Iterative Review
Author: takemo101

bytakemo101

0•

Productivity & Workflow

Automates a multi-phase quality assurance loop for documentation and configuration files to ensure technical accuracy and consistency.

This skill implements a rigorous self-improvement workflow for Claude Code, specifically designed for maintaining .opencode and .claude directories. It guides the AI through a structured cycle of planning, implementation, and a four-point review process—addressing Consistency, Completeness, Correctness, and Integration—to catch logical errors, formatting issues, and broken cross-references before changes are committed. It is an essential tool for developers who require high-precision maintenance of AI-agent definitions and complex project documentation.

Key Features

01Automated iteration management to ensure quality without infinite looping

02Structured phase-based workflow from initial planning to final commit

03Four-point review framework covering Consistency, Completeness, Correctness, and Integration

040 GitHub stars

05Severity-based issue tracking with standardized reporting for Critical, Major, and Minor issues

06Pre-commit validation checklists for repository and git integrity

Use Cases

01Maintaining complex AI agent definitions and command documentation in .claude/ directories

02Synchronizing changes across multiple JSON/YAML configuration files to prevent schema mismatches

03Large-scale documentation refactoring where cross-references and formatting must remain uniform

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add takemo101/compose-workflow iterative-review

For use in Claude.ai and ChatGPT

Download Skill