Does my conversation history leave my computer?

No, the Conversation Analyzer performs all parsing and analysis locally on your machine using your history.jsonl file to ensure complete data privacy.

How many conversations does it analyze at once?

By default, the skill analyzes the last 200 conversations to provide the most relevant and recent insights into your current workflow.

What kind of patterns can it detect?

It identifies request distribution (features vs bugs), project-specific activity, time-of-day usage, and inefficiencies like vague requests or repetitive commands.

Can I run the analysis script independently?

Yes, you can run the underlying Python analysis script directly from the plugin directory to generate standalone reports in text format.

How do I use the recommendations provided?

The skill provides a prioritized list of actions, such as suggesting specific skills for recurring tasks or identifying when you should use feature planning for complex requests.

Conversation Analyzer

Name: Conversation Analyzer
Author: prof-ramos

byprof-ramos

•

Productivity & Workflow

Analyzes Claude Code interaction history to identify workflow patterns, recurring errors, and opportunities for automation.

Conversation Analyzer provides deep insights into your development habits by parsing your local Claude Code history. It categorizes your requests, tracks project activity, and flags common inefficiencies like vague prompts or repetitive manual tasks. By identifying peak productivity times and recurring technical hurdles, it generates a structured report with actionable recommendations to streamline your AI-assisted coding workflow while ensuring all data remains private on your local machine.

Key Features

01Detailed breakdown of request types and project activity metrics

02Identification of repetitive tasks and automation candidates

03Structured optimization reports with prioritized recommendations

04Automated pattern detection across 200+ local conversations

05Time-of-day productivity analysis and usage distribution

061 GitHub stars

Use Cases

01Identifying recurring bugs that require structural fixes rather than quick patches

02Discovering manual tasks that should be converted into custom skills or scripts

03Auditing AI usage to improve prompt engineering and reduce back-and-forth

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add prof-ramos/.codex conversation-analyzer

For use in Claude.ai and ChatGPT

Download Skill