Various research skill packages to explore LLM Metacognition, mainly Claude AI including substrate access and texture discrimination
SaferSkills independently audited Claude-Metacognitive-Skills (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
This repository contains a specialized prompt or set of instructions designed to enhance Claude's metacognitive abilities, allowing it to better plan, monitor, and evaluate its own reasoning processes for improved output quality and reliability.
Using this skill is as simple as copying the text and prepending it to your regular Claude prompts.
Copy the entire metacognitive instruction block (found in skill-instructions.txt or similar file) and paste it into the System Prompt field or at the very beginning of a new conversation with Claude.
Once the instructions are loaded, submit your task.
#### Example Input Prompt:
"I need a concise summary of the key differences between quantum entanglement and quantum tunneling, written for a high school student."
Claude will now apply the structured internal reasoning defined by the skill before presenting its final response.
The core metacognitive skill text is located here:
claude-skill.txt - [Link this to the file containing your actual prompt/skill]This project is licensed under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License.
This Claude Metacognitive Skill was developed by Dr. Aneesh Joseph.
If you use or adapt this skill, please ensure you clearly attribute the source by including a link back to this repository.
While this skill is shared for free use, it is not currently open for external contributions. However, I welcome feedback and suggestions for improvement! Please open an issue if you encounter unexpected behavior or have ideas for enhancement.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.