# { "Depends": "py-genlayer:1jb45aa8ynh2a9c9xn3b7qqh8sm5q93hwfp7jqmwsfhh8jpz09h6" } from genlayer import * import json WEB_TRUNCATE = 8000 # Max chars to read from a webpage class PriceGuard(gl.Contract): """ PriceGuard — AI-Powered Price Verification on GenLayer. Users submit a product URL. GenLayer AI validators: 1. Scrape the product page to identify the item + price (gl.get_webpage) 2. Search competing sites for the same product (gl.get_webpage) 3. Compare prices and reach consensus (gl.eq_principle_strict_eq) Each check gets a verdict: GREAT_DEAL / FAIR_PRICE / OVERPRICED / SCAM_ALERT with a fair price estimate and savings percentage. """ # ═══════════════════════════════════════ # STORAGE # ═══════════════════════════════════════ total_checks: u64 # check_id -> JSON string of price check result checks: TreeMap[u64, str] # address -> total checks submitted user_checks: TreeMap[Address, u64] # Stats total_great_deals: u64 total_fair: u64 total_overpriced: u64 total_scams: u64 def __init__(self): self.total_checks = 0 self.total_great_deals = 0 self.total_fair = 0 self.total_overpriced = 0 self.total_scams = 0 # ═══════════════════════════════════════ # CORE: CHECK PRICE BY PRODUCT URL # ═══════════════════════════════════════ @gl.public.write def check_price(self, product_url: str, category: str) -> dict: """ Submit a product URL for AI price verification. The contract will: 1. Fetch the product page to identify item + listed price 2. Search competing retailers for the same product 3. Compare prices and give a verdict 4. Store the result on-chain Categories: electronics, clothing, home, beauty, sports, food, other """ if not product_url: raise Rollback("Product URL cannot be empty") check_id = self.total_checks def nondet() -> str: # Step 1: Fetch the product page product_page = gl.get_webpage(product_url, mode="text")[:WEB_TRUNCATE] # Step 2: AI identifies the product and searches for competing prices task = f"""You are PriceGuard, an expert AI price comparison analyst. TASK: Analyze the product page below, identify the product and its price, then determine if the price is fair. PRODUCT CATEGORY: {category} SOURCE URL: {product_url} PRODUCT PAGE CONTENT (delimited by <<>> and <<>>): <<>> {product_page} <<>> YOUR ANALYSIS STEPS: 1. IDENTIFY: What is the exact product name, brand, and model? 2. EXTRACT PRICE: What is the listed price on this page? (include currency) 3. MARKET KNOWLEDGE: Based on your training data, what is the typical market price range for this product or similar products? 4. COMPARE: How does the listed price compare to the typical market range? 5. CHECK FOR SCAM: Are there red flags? (unrealistically low price, suspicious seller, too-good-to-be-true deals) VERDICT RULES: - "GREAT_DEAL": Price is 20%+ below typical market value and listing seems legitimate - "FAIR_PRICE": Price is within ±15% of typical market value - "OVERPRICED": Price is 20%+ above typical market value - "SCAM_ALERT": Price is suspiciously low (80%+ below market) OR page has scam indicators (fake countdown timers, pressure tactics, no real seller info) Respond with ONLY this JSON, no prose, no code fences: {{"verdict": "GREAT_DEAL"|"FAIR_PRICE"|"OVERPRICED"|"SCAM_ALERT", "product_name": "", "listed_price": "", "estimated_fair_price": "", "savings_percent": , "reasoning": "<2-3 sentence explanation>", "price_factors": ["", ""]}}""" raw = gl.exec_prompt(task) raw = raw.replace("```json", "").replace("```", "").strip() parsed = json.loads(raw) # Normalize for cross-validator consensus normalized = { "verdict": str(parsed.get("verdict", "FAIR_PRICE")).strip().upper().replace(" ", "_"), "product_name": str(parsed.get("product_name", "Unknown Product"))[:200], "listed_price": str(parsed.get("listed_price", "N/A"))[:50], "estimated_fair_price": str(parsed.get("estimated_fair_price", "N/A"))[:50], "savings_percent": int(parsed.get("savings_percent", 0)), "reasoning": str(parsed.get("reasoning", ""))[:500], "price_factors": [str(f).strip().lower() for f in parsed.get("price_factors", [])][:5], } return json.dumps(normalized, sort_keys=True) # Reach consensus across validators result = json.loads(gl.eq_principle_strict_eq(nondet)) verdict = result.get("verdict", "FAIR_PRICE") # Build full record record = { "id": check_id, "url": product_url, "category": category, "submitter": str(gl.message.sender), "verdict": verdict, "product_name": result.get("product_name", "Unknown"), "listed_price": result.get("listed_price", "N/A"), "estimated_fair_price": result.get("estimated_fair_price", "N/A"), "savings_percent": result.get("savings_percent", 0), "reasoning": result.get("reasoning", ""), "price_factors": result.get("price_factors", []), } # Store on-chain self.checks[check_id] = json.dumps(record) self.total_checks += 1 # Update stats if verdict == "GREAT_DEAL": self.total_great_deals += 1 elif verdict == "FAIR_PRICE": self.total_fair += 1 elif verdict == "OVERPRICED": self.total_overpriced += 1 elif verdict == "SCAM_ALERT": self.total_scams += 1 # Update user check count sender = gl.message.sender current = 0 try: current = self.user_checks[sender] except: pass self.user_checks[sender] = current + 1 return result # ═══════════════════════════════════════ # CORE: MANUAL PRICE CHECK # ═══════════════════════════════════════ @gl.public.write def check_price_manual(self, product_name: str, listed_price: str, store_name: str, category: str) -> dict: """ Manually submit a product name + price for verification. AI will use its knowledge to estimate fair market value. """ if not product_name or not listed_price: raise Rollback("Product name and price are required") check_id = self.total_checks def nondet() -> str: task = f"""You are PriceGuard, an expert AI price comparison analyst. TASK: A user found a product and wants to know if the price is fair. PRODUCT: {product_name} LISTED PRICE: {listed_price} STORE: {store_name} CATEGORY: {category} YOUR ANALYSIS: 1. What is the typical retail price range for "{product_name}" or very similar products? 2. How does {listed_price} compare to the typical market range? 3. Is this store/price combination suspicious? VERDICT RULES: - "GREAT_DEAL": Price is 20%+ below typical market value - "FAIR_PRICE": Price is within ±15% of typical market value - "OVERPRICED": Price is 20%+ above typical market value - "SCAM_ALERT": Price is suspiciously low (80%+ below market) or has scam indicators Respond with ONLY this JSON, no prose, no code fences: {{"verdict": "GREAT_DEAL"|"FAIR_PRICE"|"OVERPRICED"|"SCAM_ALERT", "product_name": "{product_name}", "listed_price": "{listed_price}", "estimated_fair_price": "", "savings_percent": , "reasoning": "<2-3 sentence explanation>", "price_factors": ["", ""]}}""" raw = gl.exec_prompt(task) raw = raw.replace("```json", "").replace("```", "").strip() parsed = json.loads(raw) normalized = { "verdict": str(parsed.get("verdict", "FAIR_PRICE")).strip().upper().replace(" ", "_"), "product_name": str(parsed.get("product_name", product_name))[:200], "listed_price": str(parsed.get("listed_price", listed_price))[:50], "estimated_fair_price": str(parsed.get("estimated_fair_price", "N/A"))[:50], "savings_percent": int(parsed.get("savings_percent", 0)), "reasoning": str(parsed.get("reasoning", ""))[:500], "price_factors": [str(f).strip().lower() for f in parsed.get("price_factors", [])][:5], } return json.dumps(normalized, sort_keys=True) result = json.loads(gl.eq_principle_strict_eq(nondet)) verdict = result.get("verdict", "FAIR_PRICE") record = { "id": check_id, "url": "", "category": category, "submitter": str(gl.message.sender), "verdict": verdict, "product_name": result.get("product_name", product_name), "listed_price": result.get("listed_price", listed_price), "estimated_fair_price": result.get("estimated_fair_price", "N/A"), "savings_percent": result.get("savings_percent", 0), "reasoning": result.get("reasoning", ""), "price_factors": result.get("price_factors", []), "store_name": store_name, } self.checks[check_id] = json.dumps(record) self.total_checks += 1 if verdict == "GREAT_DEAL": self.total_great_deals += 1 elif verdict == "FAIR_PRICE": self.total_fair += 1 elif verdict == "OVERPRICED": self.total_overpriced += 1 elif verdict == "SCAM_ALERT": self.total_scams += 1 sender = gl.message.sender current = 0 try: current = self.user_checks[sender] except: pass self.user_checks[sender] = current + 1 return result # ═══════════════════════════════════════ # READ METHODS # ═══════════════════════════════════════ @gl.public.view def get_check(self, check_id: int) -> dict: """Get a single price check by ID.""" try: return json.loads(self.checks[check_id]) except: return {} @gl.public.view def get_recent_checks(self, count: int) -> list: """Get the most recent N price checks.""" results = [] start = max(0, self.total_checks - count) for i in range(start, self.total_checks): try: results.append(json.loads(self.checks[i])) except: pass return results @gl.public.view def get_stats(self) -> dict: """Get global platform statistics.""" total = max(1, self.total_checks) return { "total_checks": self.total_checks, "total_great_deals": self.total_great_deals, "total_fair": self.total_fair, "total_overpriced": self.total_overpriced, "total_scams": self.total_scams, "scam_rate": round((self.total_scams / total) * 100, 1), } @gl.public.view def get_user_stats(self, user: Address) -> dict: """Get a user's check stats.""" try: checks = self.user_checks[user] except: checks = 0 return {"checks": checks}

Hash	Type	Status		From	To	Method	GenVM Result	Consensus Result	Time
0x8a7b4bc3...8c261c56	Call	FINALIZED	IN	0xAf1D8d...a1DD27	0xfe7BbF...18cce2	check_price	ERROR	Accepted	5 months ago
0x69bc940a...e8985f25	Deploy	FINALIZED	IN	0xAf1D8d...a1DD27	0xfe7BbF...18cce2	(constructor)	SUCCESS	Accepted	5 months ago