skills.rest
samhvw8
samhvw8COMMUNITY
Explore all skills by this author →
GitHub

introspection

Analyze Claude's internal reasoning and decision patterns for transparency.

14|3|Updated Jun 24, 2021
Search Tags:#learning#decision analysis#problem-solving#meta-cognition#AI behavior#reasoning#transparency
One-click install
npx skills add https://github.com/samhvw8/dotfiles --skill introspection
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: introspection
Source: https://github.com/samhvw8/dotfiles/tree/main/dot_claude/skills/introspection
Command: npx skills add https://github.com/samhvw8/dotfiles --skill introspection

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

AI decision-making can be opaque, making it hard to understand why certain choices were made or to debug unexpected outcomes. This Skill provides a meta-cognitive analysis framework, exposing Claude's internal reasoning, decision patterns, and problem-solving approaches for transparency and continuous improvement.

Core Features & Use Cases

  • Reasoning Transparency: Explicitly reveal the "why" behind chosen approaches, alternatives considered, and decision logic.
  • Pattern Detection: Identify recurring issues, common mistakes, and successful strategies across tasks.
  • Learning Capture: Extract actionable insights from successes and failures to refine future problem-solving.
  • Use Case: Claude's solution to a coding problem didn't work as expected. Activating this Skill prompts Claude to analyze its own thought process, revealing assumptions, overlooked patterns, and potential misinterpretations, leading to a clearer path for error recovery and future prevention.

Quick Start

To trigger introspection, ask Claude to explain its reasoning.

Example:

"Explain your reasoning for choosing that approach."

"Why did you select Grep over the Task agent?"

"Analyze my decision process for the last task."