skills.rest
t6adev
t6adevCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

self-improvement

Record corrections as structured lessons in tasks/lessons.md and apply them at session start.

3|Updated Mar 19, 2026
Search Tags:#workflow#feedback#self-improvement#session-start#learning-loop#lessons-log
One-click install
npx skills add https://github.com/t6adev/claude-code-tools --skill self-improvement-t6adev
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: self-improvement
Source: https://github.com/t6adev/claude-code-tools/tree/main/tools/skills/meta/self-improvement
Command: npx skills add https://github.com/t6adev/claude-code-tools --skill self-improvement-t6adev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps you capture user corrections as structured learning lessons and reuse them to avoid repeating mistakes.

Core Features & Use Cases

  • Record mistakes and contexts in tasks/lessons.md for traceability.
  • Retrieve and apply lessons at session start to guide current tasks.
  • Evolve CLAUDE.md or rules when patterns become stable to improve responses.

Quick Start

Log the feedback in tasks/lessons.md using the prescribed format, and review existing lessons at session start to apply them.

Frequently Asked Questions about self-improvement

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I record user feedback as structured lessons in Markdown?▼

Capture user corrections and their contexts as structured entries in tasks/lessons.md. This creates a traceable log that prevents repeating the same mistakes during future tasks.

How do I apply past lessons at session start to guide current tasks?▼

Apply past lessons at session start by retrieving existing entries from tasks/lessons.md. Reviewing these structured logs informs and guides behavior for current work.

What is a learning loop for self-improvement and how does it work?▼

A learning loop for self-improvement works by capturing corrections as structured Markdown lessons and recalling them at session start. This feedback mechanism continuously guides behavior and prevents repeated mistakes.

How do I evolve CLAUDE.md rules when feedback patterns become stable?▼

Evolve CLAUDE.md rules by identifying stable patterns from logged feedback in tasks/lessons.md. Promoting recurring lessons into permanent rules improves response consistency across sessions.

Can I use this self-improvement workflow without any external dependencies?▼

You can use this self-improvement workflow without external dependencies. It operates entirely by storing and recalling logs in Markdown format within the local environment.

When should I not use a lessons log for tracking workflow feedback?▼

You should not use a lessons log for tracking workflow feedback when corrections are one-off events with no chance of recurrence. The system relies on pattern accumulation to effectively guide future behavior.

Related Skills

Explore More Personal & Entrepreneur Skills→
deathbeamdeathbeam
110

prepare-plan

Create and manage Markdown task lists in a .plans/ directory.

Community
Basic
BordaBorda
24

distill

Analyze memory, tasks, and lessons to generate structured governance proposals.

Community
Advanced
MOODMNKY-LLCMOODMNKY-LLC

session-logs

Search and analyze historical JSONL session logs with ripgrep and jq.

Community
Intermediate
marcusagmmarcusagm
4

lint-and-validate

Run linting, type checking, and security audits across Node.js/TypeScript and Python projects.

Community
Intermediate