anti-hallucination-techniques

What problem does it solve?

This Skill helps stop AI from presenting incorrect or fabricated information by forcing claims to be verified against credible sources and evidence.

Core Features & Use Cases

• Source credibility verification: Evaluate the currency, relevance, expertise, diversity, independence, and evidence quality of sources (via CREDIBLE-style checks).
• Fact-checking and validation frameworks: Apply claim validation steps (find evidence, evaluate strength, check counter-evidence, and state uncertainty).
• Hallucination pattern detection: Identify common failure modes like fake citations, incorrect technical details, and unsupported statistics.

Example use: When an AI generates an answer about an API endpoint, version, or parameter list, this Skill guides you to verify the information with authoritative documentation and to flag unsupported claims or fake references.

Quick Start

Use the anti-hallucination-techniques skill to verify whether the following AI claim is supported by credible evidence: “The Qwen Code CLI supports feature X on version Y with parameter Z…”