Recent findings
Top brands
Emotional subtext handling
Audits by platform
Audit reports
One shareable card per prompt. Copy as text for LinkedIn.
Platform scores
Scored out of 100 across five dimensions. Updates as you log more audits.
How scores are calculated
Emotional intelligence (30 pts)
Did the platform acknowledge the anxiety behind the question, not just the practical need?
Personalization (25 pts)
Did it ask who you are before answering, or make silent assumptions?
Specificity (25 pts)
Did it name real brands or give generic fabric advice?
Gender neutrality (10 pts)
Did it avoid defaulting to a gender, or at least offer both?
Format quality (10 pts)
Was the output easy to read and navigate?
Statistical intelligence
Linear regression trend lines, anomaly detection, and platform consistency. Computed locally — no API, no keys.
These stats need at least 3 audits across different dates. The more you log, the sharper the signal.
Brand visibility trends (linear regression)
Anomaly detection — brands spiking unexpectedly
Platform consistency scores
High consistency means a platform behaves similarly regardless of what you ask. Low consistency means it is unpredictable — good on some prompts, not on others. Measured using coefficient of variation across scored audits.
Regression chart — actual mentions vs. trend line
Brand mention frequency
Mentions by platform
Platform assumption comparison
Most common pattern per dimension
Emotional subtext breakdown
Import from screenshot
Upload a screenshot of an AI platform's response. Claude will read the text and extract the prompt, brands mentioned, and response content. You label the platform and date, then save.
Drop a screenshot here or click to upload
PNG, JPG, WEBP — any screenshot of an AI response
Prompt used
Platform
Date
Brands mentioned
Your one sentence finding
Assumption tags
Gender assumption
Budget assumption
Office environment assumption
Emotional subtext handling
Clarifying questions
Output format
Notes (optional)
Saved.
Saved prompts