pain

Trigger a high-priority pain signal to force system reflection and correction.

1|1|Updated Jan 22, 2026

Search Tags:#debugging #error handling #agent #reflection #intervention

Ready to run

npx skills add https://github.com/csuzngjh/principles --skill pain

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: pain
Source: https://github.com/csuzngjh/principles/tree/main/claude/skills/pain
Command: npx skills add https://github.com/csuzngjh/principles --skill pain

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill allows manual intervention when the AI agent is stuck, repeating errors, or deviating from the desired path, ensuring it corrects its course.

Core Features & Use Cases

Manual Error Injection: Directly signal a "pain point" to the system.
System Reflection Trigger: Forces the agent to re-evaluate its current state and logic.
Use Case: If the agent is stuck in a loop generating the same incorrect code, you can use this skill to inject a pain signal with your feedback, prompting it to break the loop and try a different approach.

Quick Start

Manually trigger a pain signal with the message "Agent is stuck in a loop generating incorrect code".

Frequently Asked Questions about pain

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

FAQPage Schema

How do I force an AI agent to reflect on errors and stop repeating incorrect code?▼

To force agent reflection on errors, you can manually trigger a high-priority pain signal. This injects your feedback into a designated file, prompting the agent to break its loop, re-evaluate its current logic, and correct its course during subsequent processing stages.

What is a pain signal in AI agent debugging and when should I use it?▼

A pain signal is a manual intervention mechanism used in AI agent debugging to force system reflection. You should use it when the agent is stuck, repeatedly generating the same errors, or deviating from the intended direction and needs corrective feedback.

How do I manually intervene when my agent is stuck in a loop?▼

You can manually intervene when your agent is stuck by injecting a pain signal with specific user-provided feedback. This triggers a high-priority system reflection to break the loop and try a different approach for generating code or solving the task.

Does manual error injection work for deviating AI agents without dependencies?▼

Yes, manual error injection works without external dependencies. It directly signals a pain point to the system by injecting user feedback into a designated file, which is then processed during reflection or compaction stages to correct the deviating agent.

What are the limitations of using a pain signal for agent reflection?▼

The pain signal mechanism relies on user-provided feedback to trigger reflection and does not automatically detect errors. It is a manual intervention, meaning you must actively identify when the agent is deviating or stuck and explicitly provide the reason for the pain signal.

Can I use manual intervention to correct an agent's logic during compaction stages?▼

Yes, you can use manual intervention to correct agent logic during compaction stages. The skill injects user-provided feedback as a reason for the pain signal into a designated file, which is explicitly processed during reflection or compaction phases to adjust system behavior.