Server facts
behavioral-prediction-mcp
Project
- Author
- ChainAware
- License
- Not specified
- Category
- Blockchain & Web3
Tech
- Runtime / language
- Not specified
- Transports
- stdio
- Remote available
- No
- Authentication
- Not specified
Activity
- GitHub stars
- 10
- Last commit
- Data as of
Tags
- ai
- aml
- audit
- behavioural
- fraud detection
- prediction
Overview
Wallet activity, token ownership, and contract risk form the core of ChainAware's behavioral-prediction-mcp. Its fraud analysis returns address classifications, probability scores, and forensic indicators, including money-laundering and sanctions exposure. Behavioral analysis adds likely actions, experience profiles, usage categories, and recommendations drawn from transaction history. Rug-pull analysis examines pools and contracts for signals such as ownership controls, trading restrictions, and liquidity conditions. Lenders can request wallet credit scores, while token research tools offer searchable rankings based on holder communities and individual token details with leading holders. Separate agent tools provide trust-score listings and detailed records for ERC-8004 registrations.
The project also separates immediate lookups from queued work. Fraud and behavior batch requests return a job identifier and signature; progress checks provide completion counts, and result retrieval supplies addresses for subsequent analysis rather than full wallet reports. Token auditing follows a different path: an existing completed report can be returned immediately, or a new audit is queued for later status and report retrieval. Completed audits cover ownership, liquidity, minting, transfers, approval safety, reentrancy, and honeypot findings alongside an aggregate risk assessment. Production access requires an API key, and the backend implementation remains private.
behavioral-prediction-mcp setup notes
- behavioral-prediction-mcp records stdio only; confirm a network transport upstream before planning remote access.
- behavioral-prediction-mcp has no recorded license; confirm reuse terms in its repository.