why-slow

Diagnose performance bottlenecks in Claude Code sessions from process logs and MCP health.

29|12|Updated Mar 30, 2026
One-click install
npx skills add https://github.com/matteotitta/genesys-skills --skill why-slow
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: why-slow
Source: https://github.com/matteotitta/genesys-skills/tree/main/skills/meta/session/why-slow
Command: npx skills add https://github.com/matteotitta/genesys-skills --skill why-slow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill resolves the frustration of an unresponsive or sluggish AI session by identifying the specific technical bottleneck causing the delay.

Core Features & Use Cases

  • Performance Diagnostics: Automatically checks startup logs, active MCP processes, and sync-pull debug logs to pinpoint the cause of slowness.
  • Resource Monitoring: Evaluates worktree bloat and CPU usage to determine if the environment needs cleanup or process termination.
  • Use Case: When a tool call hangs or the interface feels unresponsive, run this diagnostic to see if an MCP process is wedged or if the search domain is too large.

Quick Start

Type /why-slow in your chat interface to receive a focused report on the current session performance and recommended recovery steps.

Frequently Asked Questions about why-slow

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

FAQPage Schema
How do I troubleshoot high latency and unresponsiveness in my Claude Code session?

To troubleshoot Claude Code session latency, you need to analyze startup logs, active MCP processes, and workspace metadata to identify if an MCP process is wedged or resource usage is too high.

What is causing my tool execution to hang during an AI session?

Hanging tool execution is often caused by a wedged MCP process or an overly large search domain, which you can pinpoint by checking sync-pull debug logs and evaluating CPU usage.

How do I run a performance diagnostic on my current AI workspace?

You can run a performance diagnostic by triggering a focused report in your chat interface, which evaluates worktree bloat and active processes to recommend recovery steps.

Do I need to install special dependencies to check MCP health and workspace bloat?

You do not need special dependencies to check MCP health and workspace bloat, as the diagnostic only requires standard Unix utilities like ps, grep, tail, and git to perform non-intrusive system health checks.

What is the best way to identify if an MCP process is wedged and causing startup delays?

The best way to identify a wedged MCP process causing startup delays is to analyze active MCP processes and startup logs systematically using standard Unix command-line utilities for non-intrusive health checks.

Why does my worktree bloat cause sluggish AI session performance?

Worktree bloat causes sluggish AI session performance by expanding the search domain and consuming system resources, which can be resolved by evaluating workspace metadata and performing a cleanup.