RockHorse AI Logo

Reins

Reputation Evaluation Intelligence Network System

Monitor Share of Model, simulate RAG, and stress-test hallucinations.

Why this matters: You cannot manage what you do not measure. Reins tracks your "Share of Model" across major AI platforms and stress-tests for hallucinations, providing the executive intelligence needed to protect your brand's digital reputation in real-time.

Log in to continue

You need an account to access this module.

Brand Profile Required

You need to define your brand identity in Mare before analyzing your Share of Model.

Go to Mare

Keywords Required

You need to discover and select keywords in Colt before analyzing your Share of Model.

Go to Colt
How it works: Tracks your 'Share of Model' by polling major LLMs (ChatGPT, Claude, Gemini) with your target keywords to see how often your brand is cited as a solution.

Configuration & Scope

Query Fan-Out

Generate related queries to broaden analysis.

1 5 10

Additional Reports

Core Web Vitals

Benchmarking

Analyzing Brand...

Polling models and calculating Share of Model.

Void Check Analysis:

Share of Model

Overall Visibility

Tone Match

Profile Alignment

Sentiment Split

/ /

Pos / Neu / Neg

Multi-Model Divergence

Technical Performance Trend

Share of Model Funnel

Visibility vs. Sentiment

Entity Association Cloud

Citation Audit

Verification of key stats

Semantic Alignment

Bridge Concept usage

Run a report to see overview metrics.

Void Check:

Performance Heatmap

Keyword

Model Analysis by Keyword

Keyword ChatGPT Claude Gemini Perplexity Feedback Loop

Run a report to see detailed keyword data.

Competitive Benchmarking

Analyze competitors to see how your SoM compares.

Share of Model Trends (Brand vs. Competitors)

Run a report with Benchmarking enabled to see results.

Core Web Vitals

Technical performance audit (PageSpeed Insights).

Score
LCP
Loading
CLS
Stability
TBT
Interactivity
FCP
First Paint
SI
Visual Load
TTI
Interactive

Performance Score is weighted: TBT (30%), LCP (25%), CLS (25%), FCP (10%), SI (10%).

How it works: Map your identified "Citation Voids" against your current visibility. These are high-value targets where you can capture Share of Model by providing the missing "Strategic Asset".

Citation Void Opportunities

No voids identified yet. Use Scout in Colt to find opportunities.

How it works: Visualizes the entities defined in Mare. This confirms that the system has a structured understanding of your products, services, and people.

Knowledge Graph Tracking

Visualizing entities from Mare.

How it works: This simulator mimics how an AI (like ChatGPT or Perplexity) retrieves information from your content. By testing specific queries against your source text, you can verify if the AI is retrieving the correct 'chunks' of information to answer user questions accurately.

1. Source Content

Drag text file or paste content
Fetching content from URL...

2. Test Queries

No queries selected.

Retrieval Results

RAG Performance
Avg. Relevance
How it works: Proactively test your brand's resilience against misinformation. We generate adversarial "trap" questions and use an AI Judge to determine if the model falls for the false premise.

Red Ribbons (Active Hallucinations)

These issues need correction content.

Adversarial Configuration

1 5 10
The system will generate adversarial prompts based on your brand profile and test them against a simulated public LLM.

Test Results

Clear Historical Data?

Are you sure you want to delete all historical reports and analytics? This action cannot be undone.

Delete Report?

Are you sure you want to delete this report? This action cannot be undone.