Where is my improvement data stored?

All incident logs, session resumes, and fix registries are stored locally within the plugin's data directory, ensuring your workflow history remains private and persistent.

How do I manually trigger the improvement process?

You can initiate the protocol at any time by typing the /improve command, optionally followed by a timeframe like 'week', 'month', or 'year'.

How does the protocol detect frustration?

The skill monitors user inputs for specific trigger phrases such as 'stop,' 'again,' or 'still not working' and offers to initiate a systematic intervention.

What happens to my current work when the protocol starts?

The skill performs a Context Preservation phase, saving your current TODOs, active files, and the last user message to a resume file so you can return to work seamlessly.

Can I track the success of a fix over time?

Yes, the protocol includes an Effectiveness Tracking system that automatically checks back after 1 and 2 weeks to calculate a success rate based on pattern recurrence.

Process Improvement Protocol

Name: Process Improvement Protocol
Author: gserafini

bygserafini

0•

Productivity & Workflow

Systematically identifies and resolves workflow frustrations through root cause analysis and evidence-based fix tracking.

The Process Improvement Protocol is a specialized Claude Code skill designed to optimize developer workflows by transforming moments of friction into actionable improvements. It monitors for signs of user frustration, automatically saves session context, conducts deep root cause analysis on failure patterns, and implements persistent fixes. By tracking effectiveness over time and maintaining a 'days since last incident' metric, it ensures that your AI development environment evolves based on empirical data rather than repeated manual corrections.

Key Features

010 GitHub stars

02Evidence-based effectiveness tracking for implemented fixes

03Integrated incident logging and streak monitoring for workflow stability

04Context-aware session preservation and resume functionality

05Multi-phase root cause analysis and historical pattern comparison

06Automatic frustration detection using natural language pattern matching

Use Cases

01Reducing repetitive errors where the AI ignores specific project constraints or testing protocols

02Auditing the long-term success of custom coding standards and environment configurations

03Systematizing fixes when an AI agent fails to follow complex multi-step instructions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gserafini/process-improvement process-improvement-protocol

For use in Claude.ai and ChatGPT

Download Skill