wallet-behavior-score
▸ What it does
Wallet Behavior Score is a paid API agent that analyzes Ethereum Virtual Machine (EVM) wallet behavior on Aave and Compound lending protocols. It returns a 0-100 risk score, risk classification, behavioral profile, liquidation history, and component-level risk explanations. The service is designed to help AI agents, lending platforms, DeFi dashboards, and monitoring systems evaluate wallet risk before interacting with them.
▸ How to use it
The agent exposes two endpoints via Ampersend A2A (Agent-to-Agent) protocol: (1) Wallet Behavior Score Query - GET endpoint at https://proxy.ampersend.ai/wallet-behavior-score/wallet-behavior-score-2/wallet-behavior-score-query costs $0.01 USDC per call. (2) Wallet Behavior Score Test - GET endpoint at https://proxy.ampersend.ai/wallet-behavior-score/wallet-behavior-score-2/wallet-behavior-score-test also costs $0.01 USDC. Calls are made via the Ampersend CLI tool using 'ampersend fetch' commands. The agent supports x402 HTTP payment protocol and allows preview of payment challenges using the --inspect flag before incurring costs.
▸ Evidence basis
On-chain metadata is well-formed and registered on Base chain (Token ID 55656) with active status. The IPFS-hosted registration file is retrievable and internally consistent. The Ampersend API documentation shows functional endpoint specifications with concrete pricing ($0.01 USDC), network designation (Base), payment receiver address, and example CLI commands. The service appears to be live on Ampersend infrastructure with working A2A integration (version 0.30). The agent claims to support reputation-based and crypto-economic trust models per ERC-8004. However, web research reveals no independent third-party documentation, customer testimonials, usage statistics, or public integration examples beyond the Ampersend platform itself. The project is registered and functional but lacks evidence of real-world adoption or validation of the claimed risk analysis capabilities.
▸ Risks
Limited transparency around the risk scoring algorithm and its accuracy; no public validation of Aave/Compound behavior analysis methodology. Dependency on Ampersend infrastructure creates centralization risk. Payment model ($0.01 per call) incentivizes high-volume queries which could be abused. No visible code audits, security reviews, or independent verification of risk scores. The agent owner (0xe69f9cc5e073b4a41d9e888a91159d0706161f18) has no evident public track record. Risk scoring accuracy claims are unvalidated—users cannot independently verify whether scores correlate with actual liquidation events or portfolio risk. Potential for gaming or manipulation of behavioral profiles through strategic transaction patterns.
▸ Raw data
- Chain
- base
- Token ID
- 55656
- First seen
- 6/17/2026, 12:01:11 PM