Self-Analyze

Analyze outputs to identify root causes and create improvement plans.

Updated Jul 24, 2025
One-click install
npx skills add https://github.com/AX661s/openclaw-metacog-template --skill self-analyze
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Self-Analyze
Source: https://github.com/AX661s/openclaw-metacog-template/tree/main/skills/self-analyze
Command: npx skills add https://github.com/AX661s/openclaw-metacog-template --skill self-analyze

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables deliberate, structured analysis of your own outputs to identify and implement improvements.

Core Features & Use Cases

  • Provides a repeatable framework for evidence gathering, root-cause analysis, and pattern recognition to reduce repeated mistakes.
  • Guides creating improvement proposals, modification logs, and verification plans to close the loop on learning.
  • Supports periodic heartbeats or triggered reviews for ongoing quality assurance across tasks.

Quick Start

After analyzing outputs, run the Self-Analyze skill to start a guided review and produce a structured improvement plan.

Frequently Asked Questions about Self-Analyze

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

FAQPage Schema
What is structured self-analysis for AI output improvement?

Structured self-analysis is a repeatable framework to identify and fix quality issues in outputs through evidence gathering, root-cause analysis, and pattern recognition to reduce repeated mistakes. It closes the learning loop by producing modification logs and verification plans.

How do I perform root-cause analysis on AI errors?

Perform root-cause analysis on AI errors by running a structured post-output review to gather data, analyze patterns, and generate an improvement proposal. This process logs modifications and creates a verification plan to prevent recurring mistakes.

Can I schedule automated self-analysis for ongoing quality assurance?

Yes, you can schedule automated self-analysis for ongoing quality assurance across tasks. The framework supports periodic heartbeats or triggered reviews to continuously identify output issues and apply structured modifications.

What's the best way to track and correct recurring output errors?

The best way to track and correct recurring output errors is by maintaining a modification log with a verification plan. This structured approach uses pattern analysis to identify root causes and implement targeted improvements.

Do I need a data-gathering protocol before starting meta-cognition reviews?

Yes, a data-gathering protocol is required before starting meta-cognition reviews. This protocol provides the necessary evidence and output history to accurately perform root-cause analysis and identify recurring quality patterns.