dag-hallucination-detector

Verify citations and claims in DAG outputs to detect hallucinations.

10|1|Updated Mar 17, 2026
One-click install
npx skills add https://github.com/curiositech/windags-skills --skill dag-hallucination-detector
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dag-hallucination-detector
Source: https://github.com/curiositech/windags-skills/tree/main/skills/dag-hallucination-detector
Command: npx skills add https://github.com/curiositech/windags-skills --skill dag-hallucination-detector

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Detects fabricated content, false citations, and unverifiable claims in agent outputs, helping prevent misinformation from propagating through DAG pipelines.

Core Features & Use Cases

  • Verification of quotes and citations across outputs.
  • Consistency and temporal integrity checks to catch internal contradictions.
  • Integration with adjacent skills (dag-output-validator, dag-confidence-scorer) for a robust validation workflow.

Quick Start

Provide a real-time hallucination assessment for a given DAG output.

Frequently Asked Questions about dag-hallucination-detector

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

FAQPage Schema
How do I detect hallucinations and verify fabricated claims in AI agent outputs?

Hallucination detection in AI agent outputs identifies fabricated content, false citations, and unverifiable claims by verifying quotes and performing factual consistency checks to prevent misinformation.

What causes AI pipelines to output false citations and internal contradictions?

AI pipelines output false citations and contradictions when lacking traceable fact-checking. Consistency and temporal integrity analysis catches these internal contradictions by verifying claims against source data.

How do I perform citation verification and fact-checking for DAG outputs?

Citation verification for DAG outputs is performed through a configurable workflow that checks factual claims, verifies quotes across outputs, and detects hallucination patterns within AI pipeline results.

Does this hallucination detection approach work within existing AI validation workflows?

This hallucination detection integrates with existing AI validation workflows by connecting with dag-output-validator and dag-confidence-scorer to provide a robust, traceable verification pipeline.

What is the best way to stop misinformation from propagating through automated AI pipelines?

The best way to stop misinformation in automated AI pipelines is implementing real-time hallucination assessments that verify citations and check factual claims before outputs are finalized.