The problem MiddleKid solves

MiddleKid makes DeFi safer and more accessible by protecting users from scams and simplifying complex blockchain analysis.

🔒 Scam Protection
DeFi users lose millions to honeypot tokens and rug pulls. MiddleKid's multi-source Researcher combines GoPlusLabs security scans + DEX Screener market data to detect:

• Honeypot tokens (can't sell after buying)
• High buy/sell taxes (>10%)
• Hidden owners & malicious contracts
• No liquidity or fake volume
  Before buying any token on Base, users can instantly check if it's legitimate or a scam - preventing costly mistakes.

💰 Simplified Wallet Analysis
Instead of navigating multiple block explorers, users can:

View all Base holdings in one clean interface
Track token balances with real-time USD values
Monitor DeFi positions and health scores
🎯 User-Friendly Experience
No technical knowledge required - Just paste a token address
Clear risk scores (0-100) with actionable flags
Multi-source verification - Not relying on single data provider
🚀 Future Features (Planned)
Translator: Decode complex transactions into plain English
Time Traveller: Historical price contracts for strategic trading
AI Chat: Get instant answers about tokens, risks, and DeFi strategies
MiddleKid acts as your "sidekick" in DeFi - making the Base ecosystem safer, easier to navigate, and more accessible to everyone, from beginners to advanced traders.

Challenges I ran into

Challenge 1: Risk Scoring Accuracy
The Problem: Initial implementation used only Alchemy's getAssetTransfers API to count holders and transfers. Both legitimate tokens (SHIBA, Polkadot) and scam tokens showed similar transfer activity - making the system useless for scam detection.

Solution: Pivoted to a multi-source approach:

Integrated GoPlusLabs API for security scans (honeypot detection, tax analysis)
Added DEX Screener API for market legitimacy (liquidity, volume)
Implemented weighted scoring (60% security, 30% market, 10% metadata)
This reduced false positives dramatically while catching actual scams.

Challenge 2: Balancing Scoring Penalties
The Problem: Many legitimate tokens are "mintable" or have certain owner privileges, causing high risk scores for safe tokens like Polkadot (scored 12/100 when it should be ~3).

Solution:

Reduced mintable penalty from 20 → 5 points (common in legit tokens)
Increased "no market data" penalty from 30 → 50 points (high risk indicator)
Added granular liquidity thresholds (<$1K = 40pts, <$10K = 25pts, etc.)
🎨 Challenge 3: Z-Index Layering Chaos
The Problem: Navigation tabs appeared behind the wallet connection modal - but only after the modal opened. Tried increasing navigation z-index to 9999, still didn't work!

Root Cause: The wallet button wrapper had z-index: 101, creating a stacking context that isolated its children from the global z-index hierarchy.

Solution: Set top bar elements (logo, title, wallet) to z-index: 10000 and navigation to z-index: 100. Added click-outside-to-close for better UX.

Challenge 4: State Management
The Problem: Researcher page required hard refresh between searches - results from previous search persisted.

Solution: Implemented proper state clearing with a unique key prop on ResearchResults component based on search query, forcing React to unmount/remount on new searches.

Challenge 5: API Rate Limiting & Reliability
The Problem: Free APIs (GoPlusLabs, DEX Screener) don't always have data for every token.

Solution: Built fallback scoring system - if an API fails or returns no data, assign conservative penalties instead of crashing. Added detailed console logging for debugging.