Framework Critical Thinking Framework (FCT)

Overview

Critical Thinking Framework (FCT) provides architectural components for building AI agents with metacognition and self-correction capabilities. This framework integrates state-of-the-art techniques such as Chain-of-Thought, Tree-of-Thoughts, and self-verification to produce more reliable and transparent reasoning.

When to use this skill:

• Building AI agents for complex tasks requiring multi-step reasoning
• Implementing self-correction and error detection in agent workflows
• Selecting the appropriate reasoning method (CoT vs ToT vs GoT) based on task characteristics
• Adding metacognitive monitoring to improve reliability
• Reducing hallucination and reasoning errors in AI outputs

Triggers: Use this skill when the user requests help with:

• "build an agent that can self-correct"
• "implement better reasoning"
• "reduce hallucination in AI system"
• "choose between CoT, ToT, or other reasoning methods"
• "add error detection to agent workflow"