mistake-tracking

Community

Log, analyze, and prevent recurring mistakes.

Authormshuffett
Version1.0.0
Installs0

System Documentation

What problem does it solve?

This Skill provides a robust system for identifying, logging, analyzing, and ultimately preventing recurring mistakes or near-misses within an AI agent's workflow, ensuring continuous improvement and system reliability.

Core Features & Use Cases

  • Mistake Logging: Records mistakes with detailed context (timestamp, type, scope, notes, actions taken).
  • Root Cause Analysis: Facilitates understanding why a mistake occurred and why the system didn't prevent it.
  • Guardrail Management: Promotes or demotes system-level guardrails based on mistake frequency and patterns.
  • Testing & Validation: Includes mechanisms to test fixes and verify guardrail effectiveness.
  • Use Case: After an agent makes a critical error, this Skill logs the event, helps determine if a new guardrail is needed or an existing one needs strengthening, and provides a framework to test the fix before it's deployed.

Quick Start

Log a new mistake by running the log-mistake script with a JSON event detailing the error.

Dependency Matrix

Required Modules

None required

Components

scriptsreferencesassets

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: mistake-tracking
Download link: https://github.com/mshuffett/dotfiles/archive/main.zip#mistake-tracking

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.