Prompt Coach

Analyze Claude Code session logs to quantify prompt quality and tool usage.

Updated Jan 4, 2026

Search Tags:#productivity #MCP #Claude Code #log-analysis #prompt-analysis #token-usage #tool-usage

Ready to run

npx skills add https://github.com/refly-ai/skill-to-workflow --skill prompt-coach-refly-ai

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Prompt Coach
Source: https://github.com/refly-ai/skill-to-workflow/tree/main/skills-source/zijiebijiben/claude-code-prompt-coach-skill-main
Command: npx skills add https://github.com/refly-ai/skill-to-workflow --skill prompt-coach-refly-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps developers transform Claude Code session data into actionable improvements for prompt quality, tool usage, and overall coding efficiency.

Core Features & Use Cases

Usage analytics: derive metrics from local session logs to reveal patterns in prompts, tokens, and tool usage.
Workflow optimization: surface recommendations to reduce context switching, improve prompts, and increase MCP adoption.
Report-ready outputs: generate structured insights suitable for dashboards and performance reviews.

Quick Start

To get started, ask Claude Code to analyze your local Claude Code usage across all projects and review the resulting comprehensive report.

Frequently Asked Questions about Prompt Coach

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

FAQPage Schema

How do I analyze Claude Code session logs to improve prompt quality?▼

Analyzing Claude Code session logs requires parsing local data from ~/.claude/projects/ to quantify prompt quality, tool usage, and productivity. This process generates context-aware insights and actionable recommendations to improve coding efficiency.

What is the best way to track token usage and tool usage in Claude Code?▼

Tracking token usage and tool usage in Claude Code involves extracting metrics from local session logs. This approach deduplicates token usage, models distribution, and produces structured reports suitable for personal dashboards.

How do I generate report-ready insights from Claude Code usage data?▼

Generating report-ready insights from Claude Code usage data involves processing local session logs to surface workflow optimization recommendations. It creates structured outputs that support personal dashboards and performance reviews.

Can I use local Claude Code logs to increase MCP adoption and reduce context switching?▼

Local Claude Code logs can be analyzed to increase MCP adoption and reduce context switching. The analysis surfaces specific workflow optimization recommendations by evaluating prompt patterns and tool usage metrics.

Does analyzing Claude Code usage require any external dependencies or database setup?▼

Analyzing Claude Code usage does not require external dependencies or database setup. The skill operates directly on local session logs stored in the ~/.claude/projects/ directory to derive usage metrics and insights.