experimental-self-prompting

Generate context-free prompts to reframe stalled analysis.

8|1|Updated Jan 24, 2026

Search Tags:#code review #verification #analysis #bias mitigation #self-prompting #context-free

Ready to run

npx skills add https://github.com/bordenet/superpowers-plus --skill experimental-self-prompting

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: experimental-self-prompting
Source: https://github.com/bordenet/superpowers-plus/tree/main/skills/experimental/experimental-self-prompting
Command: npx skills add https://github.com/bordenet/superpowers-plus --skill experimental-self-prompting

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill addresses stalled or biased analysis caused by context‑window pollution, anchoring bias, or over‑reliance on external models, providing a fresh, context‑free perspective.

Core Features & Use Cases

Self‑generated prompts: Writes comprehensive, context‑free prompts that any engineer could understand without prior knowledge.
Verification workflow: Mandates manual confirmation of each finding to prevent hallucinations.
Safety constraints: Explicitly blocks usage with external models and highlights a ~20 % false‑positive rate.
Applicable scenarios: Complex system reviews, adversarial analysis, pre‑commit checks, and independent verification of external model outputs.

Quick Start

Use the experimental-self-prompting skill to generate a context‑free analysis prompt for the target system.

Frequently Asked Questions about experimental-self-prompting

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

FAQPage Schema

How do I reduce bias and hallucination risk during complex code reviews?▼

This approach creates context-free prompts to reframe stalled analysis, mitigating anchoring bias and context-window pollution. It provides a fresh perspective for complex code reviews without relying on external models.

What is a context-free prompt for adversarial analysis?▼

A context-free prompt for adversarial analysis is a self-crafted instruction set understandable without prior conversation knowledge. It reframes target system assessments independently to prevent context-window pollution and bias.

Can I use self-prompting for pre-commit verification checks?▼

Yes, self-prompting applies to pre-commit verification by generating independent analysis prompts for complex systems. It mandates manual confirmation of each finding due to a roughly 20% false-positive rate.

Do I need an external model to generate unbiased analysis prompts?▼

No, external models are explicitly blocked to prevent hallucinations. The skill self-generates comprehensive prompts internally to maintain a safe, context-free verification workflow for the target system.

What are the limitations of using self-crafted prompts for code review?▼

Limitations include a roughly 20% false-positive rate, requiring manual verification of all findings before action. It is designed for stalled analysis and should not replace standard review processes without independent confirmation.