ReviewerAgent (Anti-Hallucination)

Verify academic citations against Semantic Scholar, CrossRef, and Europe PMC.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/tajo9128/BioDockify-pharma-research-ai --skill revieweragent-anti-hallucination
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ReviewerAgent (Anti-Hallucination)
Source: https://github.com/tajo9128/BioDockify-pharma-research-ai/tree/main/agent_zero/skills/reviewer_agent
Command: npx skills add https://github.com/tajo9128/BioDockify-pharma-research-ai --skill revieweragent-anti-hallucination

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides an automated peer-review layer for AI-generated research content, ensuring citations map to real papers to prevent hallucinations.

Core Features & Use Cases

  • Citation Extraction: Identifies [Author, Year] citations and DOIs from text.
  • Cross-Database Verification: Validates citations against Semantic Scholar, CrossRef, and Europe PMC.
  • Integrity Scoring: Produces an overall reliability score for the text.
  • Evidence Mapping: Locates the closest real paper for suspicious citations.
  • Usage Context: When drafting manuscripts or summaries, run this skill to validate references before submission.

Quick Start

Use the ReviewerAgent to verify citations by calling verify_citations on your draft text.

Frequently Asked Questions about ReviewerAgent (Anti-Hallucination)

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

FAQPage Schema
How do I verify academic citations to prevent AI hallucinations in research drafts?

You can verify citations for AI-generated text by extracting [Author, Year] references and DOIs, then cross-checking them against Semantic Scholar, CrossRef, and Europe PMC to ensure they map to real papers and receive an integrity score.

How does automated peer-review score citation integrity?

Citation integrity scoring works by cross-referencing extracted citations against Semantic Scholar, CrossRef, and Europe PMC, producing a reliability score and mapping suspicious references to the closest real papers found in those databases.

Do I need Python and network access to use automated citation verification?

Yes, automated citation verification requires a Python-based reviewer component with network access to external academic databases and a wrapper that exposes the verify_citations method to function properly during drafting and submission workflows.

What happens when citation verification finds suspicious or hallucinated references?

When suspicious citations are detected, the evidence mapping feature locates the closest real paper within Semantic Scholar, CrossRef, or Europe PMC, allowing you to correct hallucinated references before finalizing your research draft.

Can I check DOIs and [Author, Year] citations against multiple academic databases simultaneously?

Yes, automated citation verification cross-references DOIs and [Author, Year] citations against Semantic Scholar, CrossRef, and Europe PMC simultaneously, extracting references from your text and validating them across all three databases for comprehensive coverage.