self-reflection

Verify factual claims in user outputs using parallel isolated subagents.

Updated Apr 17, 2026
One-click install
npx skills add https://github.com/jiesou/dotfiles --skill self-reflection-jiesou
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-reflection
Source: https://github.com/jiesou/dotfiles/tree/main/exact_dot_agents/exact_skills/exact_0research/exact_self-reflection
Command: npx skills add https://github.com/jiesou/dotfiles --skill self-reflection-jiesou

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill helps users verify the accuracy of their problem-solving outputs by independently auditing factual claims.

Core Features & Use Cases

  • Factual Claim Verification: Identifies and verifies factual claims in user outputs.
  • Isolated Context: Subagents operate in isolated contexts to ensure unbiased verification.
  • Parallel Processing: Subagents verify claims in parallel for efficiency.
  • Iterative Loop: Offers an iterative loop for continuous verification and correction until the output is accurate.
  • Use Case: Ideal for verifying the accuracy of complex problem-solving outputs, such as research findings or business analysis.

Quick Start

Run the self-reflection skill to verify the accuracy of your latest research findings.

Frequently Asked Questions about self-reflection

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

FAQPage Schema
How do I verify the factual accuracy of my research findings?

To verify factual accuracy, this skill identifies claims in your research findings and executes isolated subagents in parallel to independently audit them. It runs an iterative loop for continuous verification and correction until the output is accurate.

How does isolated subagent verification work for problem-solving outputs?

Isolated subagent verification works by deploying independent agents in parallel to check factual claims without shared context bias. This ensures unbiased problem-solving verification before any iterative corrections are applied to the output.

Do I need external dependencies to run iterative fact-checking on my outputs?

No external dependencies are required to run iterative fact-checking. The verification process operates independently using internal subagents to identify and correct factual claims in your problem-solving outputs.

Can I use this for verifying complex business analysis outputs?

Yes, you can use this for verifying complex business analysis outputs. It is specifically designed to audit factual claims in complex problem-solving results through parallel independent subagents and an iterative correction loop.

What is the best way to continuously correct inaccurate factual claims in generated text?

The best way to continuously correct inaccurate factual claims is using an iterative verification loop. Independent subagents identify and audit claims in parallel, repeating the process until the generated text reaches complete factual accuracy.