Prompt Details
Model
(gemini-3.5-flash)
Token size
662
Example input
[raw_e_commerce_data]: 14:22:05 POST /api/v1/coupon/apply 422 (Unprocessed Entity) - "coupon_expired". User ID 8842 stayed on /checkout page for 180 seconds scrolling back and forth near div.coupon-wrapper.
Example output
{
"target_url_pattern": "/checkout",
"trigger": {
"type": "api_error",
"condition_meta": "422_coupon_expired"
},
"ui_mutations": [
{
"element_selector": "div.coupon-wrapper",
"action": "inject_html",
"payload": {
"content_html": "<div class='discount-alert'>⚠️ This coupon has expired, but we've applied an automatic 5% off code <strong>WELCOME5</strong> just for you!</div>",
"attributes": {},
"css_override": {
"border": "2px solid #ef4444",
"background-color": "#fef2f2",
"padding": "12px",
"border-radius": "6px"
},
"delay_ms": 200
}
}
],
"analytics_tracking": {
"root_cause_analysis": "User got stuck after a 422 error from an expired coupon code, halting progress.",
"severity_score": "high",
"estimated_recovery_probability": 0.85
}
}
By purchasing this prompt, you agree to our terms of service
GEMINI-3.5-FLASH
The concept: A system prompt that tightly integrates analytics with data structure. An entrepreneur feeds the AI a raw error log exported from the website or a dataset on customer behavior (e.g., who abandoned their cart and at what stage).
What the AI does: It parses this data strictly according to established rules and outputs—not just text—but a perfectly valid JSON configuration file for UI/UX elements. This file specifies exactly which website sections to hide and which prompts to disp
...more
Added 2 weeks ago
