In the Bonafide platform, the Curator module serves as your brand's control center for managing the information Large Language Models (LLMs) consume about your destination or property. As you navigate the hundreds of questions (prompts) users are asking AI about your brand, you will notice they are organized by their "Verification Status".
These statuses—color-coded for easy identification—tell you exactly where an answer came from and whether a human has approved it. Understanding these statuses is the first step to effectively prioritizing your curation efforts and taking control of your AI narrative.
1. System of Record (Light Green)
When a prompt is marked as System of Record (represented by the color light green), it means that Bonafide’s web crawler successfully found an answer to the question directly on your official website.
Because your website acts as the primary source of truth, Bonafide automatically extracts this content and drafts an "Official Response". However, while this information comes from your site, the "System of Record" status indicates that a human user has not yet manually reviewed or confirmed the drafted answer.
2. Unverified (Yellow)
The Unverified status (represented by gray in the bar graphs, and often a yellow row in the detail view) indicates that an answer has not been confirmed by a human. More importantly, it usually means that a clear answer was not found on your official website during the crawl. Unverified questions represent your brand's largest content gaps, as they highlight what users are asking AI, but your website fails to answer.
When dealing with Unverified prompts, you will typically encounter two scenarios in the "Official Response" column:
- Blank Responses: If the response is blank, it means the LLMs could not find a definitive consensus anywhere on the internet to answer the question.
- AI-Recommended Responses (LLM Consensus): Sometimes, an Unverified row will already have text in it. This means that while the answer wasn't on your website, a majority of the LLMs agreed on an answer from other open-web sources, and the system is providing that consensus as a recommended starting point for you to review.
Crucial Note on Unverified Data: Any prompts that are left in the "Unverified" status will not be included in your final FAQ markdown files (your LLM knowledge base). Until a human verifies the information, it is held back from distribution.
3. Verified (Dark Green)
The Verified status (represented by dark green) is the ultimate goal for your prompts. This status signifies that a human user has physically gone into the Curator system, reviewed the prompt, edited or accepted the Official Response, and actively clicked the "Verified" button.
By changing a status to Verified, you are establishing human-in-the-loop confirmation. You are effectively telling the LLMs, "I am the authoritative source, and this is the correct answer". Whether you are approving a "System of Record" answer or filling in a blank "Unverified" gap, verifying the response ensures it is officially owned by your brand and ready to be packaged for the AI ecosystem.