How does it help with token optimization?

The skill includes a 'calculate-context-load' operation that analyzes your project structure to identify token-heavy areas and optimization opportunities for better context management.

Is this skill compatible with all Claude Code projects?

It is specifically designed for projects following the .claude/ directory structure and FABER-agent standards, but it can be used to audit any compatible directory for structural patterns.

Does Project Analyzer modify my project files?

No, Project Analyzer is strictly a read-only tool. It executes bash scripts to inspect the project structure and patterns without making any changes to your source code.

What kind of architectural errors does this skill detect?

It identifies anti-patterns like project-specific directors, missing workflow logs, improper command routing, and Manager-as-Skill configurations that can break agent workflows.

Project Analyzer

Name: Project Analyzer
Author: fractary

$fractary$ byfractary

•

分析与监控

Audits Claude Code projects to detect architectural anti-patterns and optimize structural compliance through automated detection scripts.

Project Analyzer is a diagnostic utility designed to maintain the architectural integrity of Claude Code projects. It performs non-destructive, read-only audits by executing specialized bash scripts that identify common anti-patterns, such as project-specific managers or directors that bypass standard workflows. By evaluating project structure, verifying workflow logging compliance, and calculating context token loads, this skill helps developers maintain scalable, efficient, and standardized AI agent environments while preventing common implementation errors.

主要功能

01Standardized JSON output for project compliance scores and structural counts

02Detection of project-specific director and manager anti-patterns

032 GitHub stars

04Identification of direct skill-to-command routing violations

05Automated architectural auditing for Claude Code project structures

06Context token load calculation and optimization reporting

使用场景

01Optimizing AI context window usage by identifying high-token-load files and structures

02Refactoring legacy Claude Code projects to eliminate architectural debt and improper agent routing

03Performing a pre-migration audit to ensure project compatibility with FABER-agent standards

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add fractary/claude-plugins project-analyzer

For use in Claude.ai and ChatGPT

Download Skill