agent-complexity-optimizer

This Skill addresses the challenge developers face in identifying hidden performance bottlenecks during coding. In practice, teams often encounter severe performance degradation under high concurrency due to inefficient patterns like nested loops, N+1 queries, and sorting within loops, which are easily missed during manual code reviews. This tool automatically scans the entire codebase, precisely pinpoints O(n²) hotspots and code segments with excessive complexity, and provides structured remediation suggestions, all without modifying any files.

Usage is straightforward. Simply execute npx agent-complexity-optimizer in your terminal. It will automatically detect the AI coding agents (such as Claude Code, Cursor, Windsurf, etc.) installed in your current environment and write the correct configuration format. For Claude Code users, it can also be installed directly via the plugin marketplace: /plugin marketplace add sebastianbreguel/agent-complexity-optimizer. You can then trigger an analysis using natural language commands like 'scan project complexity', which outputs a comprehensive report including severity levels, file locations, and optimization suggestions.

It is particularly well-suited for development teams or individual developers who need to continuously optimize code quality. This is especially true for teams already using multiple AI coding agents (like Claude Code, Codex, Copilot, etc.), as they can obtain complexity analysis through a single unified tool, avoiding the need to switch between different platforms. For backend developers, full-stack engineers, and agent developers, this tool can significantly reduce the risk of performance regressions during code reviews.

It is recommended to run this tool regularly during project iterations, such as before each commit or integrated into CI/CD pipelines. Please note that it primarily targets algorithmic complexity patterns (nested loops, N+1 queries, etc.) and is not a general-purpose code quality checker. For Python, it uses AST parsing; for other languages, it employs regex matching, which may result in a small number of false positives or false negatives. Manual verification alongside code review results is advised.