Analyze your Claude Code session logs to improve prompt quality, optimize tool usage, and become a better AI-native engineer.
Prompt Coach is an AI skill designed to help developers improve their prompt engineering and AI-assisted coding workflows by analyzing local Claude Code session logs. By examining the patterns in your interactions with Claude Code, it identifies areas where prompts can be more effective, tools are underutilized, and coding efficiency can be enhanced. This skill addresses the challenge of optimizing AI-driven development processes by providing actionable insights from historical session data, enabling developers to refine their approach and maximize the value of AI tools in their workflow.
The skill offers a range of capabilities, including evaluating prompt quality, measuring tool usage, tracking session efficiency, identifying peak productivity hours, highlighting frequently modified files, assessing context switching overhead, and analyzing common error patterns and recovery strategies. Users can perform a general analysis across all projects, inspect specific projects, or list available projects before selecting one to analyze. Reports include detailed metrics on token usage, prompt clarity, tool adoption, session efficiency, productivity patterns, file modification hotspots, and error handling behaviors, giving developers a comprehensive understanding of their coding habits and AI tool interactions.
Prompt Coach is particularly useful for AI-native engineers, prompt engineers, and developers looking to integrate AI more effectively into their coding workflow. It helps users learn from past sessions, optimize their prompt formulation, improve productivity, and make data-driven decisions on AI tool usage. Targeted at individuals who work locally with Claude Code session logs, this skill enables a personalized approach to improving AI-assisted development practices and achieving more efficient, structured, and productive coding sessions.
To use Prompt Coach, ask for a general analysis or specify a project path to analyze the local Claude Code session logs stored in '~/.claude/projects/'.
No, Prompt Coach can only access and analyze session logs stored locally on the machine where it is installed.
You can generate general analyses covering token usage, prompt quality, tool usage, session efficiency, productivity times, file hotspots, error patterns, and context switching, or focus on specific projects for targeted insights.
Yes, providing the full path to the project directory ensures that Prompt Coach analyzes the correct session logs.
You need Claude Code installed with session logs available locally under '~/.claude/projects/'. No other specific requirements are mentioned.
Quick Setup:
.claude/skills/Repository
bear2u/my-skills