Where does the data for these analytics come from?

The skill pulls data from project memory files like LEARNED.md and CLAUDE.md, as well as your local Git history and diff stats to build a complete picture of your activity.

Does it track my productivity over time?

Yes, the skill provides metrics for your last 10 sessions, including average session length, edit volume, and capture rate for new learnings to help you see progress.

What are 'stale learnings'?

Stale learnings are patterns or rules previously captured that haven't been applied in a significant amount of time, indicating they may be outdated and should be removed from your context.

Can this skill help me improve my AI coding efficiency?

Yes, it identifies 'hot' areas with high correction rates, suggesting where you should implement permanent rules to prevent the AI from making the same mistakes repeatedly.

How does the insights skill calculate corrections?

It scans session history and learning logs for instances where you redirected or overrode the agent's output, helping identify friction points in your workflow.

AI Workflow Insights

Name: AI Workflow Insights
Author: rohitg00

byrohitg00

•

1,442

•

Analytics & Monitoring

Generates detailed productivity metrics and learning patterns from session history to optimize AI coding efficiency.

The Insights skill provides a data-driven dashboard for your Claude Code sessions by analyzing project memory and Git activity to surface actionable intelligence. It tracks session duration, file modifications, and correction rates while identifying which learned patterns are most effective and which recurring errors require new permanent rules. By visualizing friction points through correction heatmaps and identifying stale instructions, it helps developers refine their 80/20 AI coding ratio and maintain a clean, efficient project context.

Key Features

01Productivity trend tracking across multiple sessions to monitor workflow improvement

021,442 GitHub stars

03Comprehensive session summaries tracking duration, edits, and self-corrections

04Correction heatmaps that highlight recurring friction points in testing, navigation, and quality

05Learning analytics that categorize and monitor the application frequency of captured patterns

06Actionable guardrails that suggest permanent rules for high-frequency correction areas

Use Cases

01Identifying and removing 'stale' project rules that are no longer applied or relevant

02Reviewing daily productivity and AI interaction efficiency after completing a development task

03Converting recurring AI mistakes into permanent project instructions via correction trend analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rohitg00/pro-workflow insights

For use in Claude.ai and ChatGPT

Download Skill