anti-hallucination

Enforces strict verification of AI-generated code, facts, and references before outputting them.

34|10|Updated Mar 12, 2026

Search Tags:#fact-checking #verification #code execution #reliability #accuracy #anti-hallucination

Ready to run

npx skills add https://github.com/dariia-m/my_claude_skills --skill anti-hallucination

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: anti-hallucination
Source: https://github.com/dariia-m/my_claude_skills/tree/main/dont-lie
Command: npx skills add https://github.com/dariia-m/my_claude_skills --skill anti-hallucination

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents the AI from fabricating information, ensuring all outputs are factually accurate and based on verified data, code, or documentation.

Core Features & Use Cases

Information Verification: Guarantees that any facts, code, file paths, or function arguments are checked before being used.
Code Execution: Mandates that all generated code is run and errors are fixed, rather than presented as untested.
Use Case: When asked to write code to process a specific file, this skill ensures the AI first checks if the file exists and then runs the code, fixing any errors before presenting the solution.

Quick Start

Use the anti-hallucination skill to verify all information before responding.

Frequently Asked Questions about anti-hallucination

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

FAQPage Schema

How do I stop AI from hallucinating fake file paths and function arguments?▼

To stop AI hallucinations of fake file paths and function arguments, you need a strict verification protocol that checks data existence and reads documentation before generating responses. This guarantees outputs rely on verified information.

How to ensure generated code is actually run and tested before presenting it?▼

To ensure generated code is actually run and tested, apply an anti-hallucination protocol that mandates code execution and error fixing prior to output. This prevents presenting untested code by requiring errors to be resolved during generation.

What is AI hallucination prevention and how does fact-checking work?▼

AI hallucination prevention is a strict fact-checking protocol that verifies information, code, and data before output. It works by requiring adherence to rules for reading documentation, executing code, and explicitly stating uncertainty to prevent fabricated content.

Can I enforce explicit uncertainty statements when the AI lacks verified data?▼

Yes, you can enforce explicit uncertainty statements by applying a strict anti-hallucination protocol that requires the AI to state uncertainty explicitly when it lacks verified data. This prevents fabricated facts by mandating honesty about missing information.

Does fact-checking AI output slow down code generation workflows?▼

Fact-checking AI output adds a verification layer to code generation by requiring file existence checks, code execution, and error fixing before delivery. This ensures reliability and accuracy, preventing untested code from reaching production.