ZKSentinel

ZKSentinel: Secure, Verifiable Autonomous Trading Overview ZKSentinel is an autonomous AI trading agent designed for the Agentic Web (2026), bridging the gap between sophisticated momentum strategies and on-chain trust. Built for the Lablab.ai AI Trading Agents Hackathon, it addresses the "Black Box" problem of automated trading by implementing the ERC-8004 standard for verifiable intent. Technical Strategy The agent utilizes the Volume-Confirmed Momentum Strategy, a three-pillar voting system that requires confluence before execution: Trend Analysis: A 9-period EMA filters primary price direction. Volume Integrity: On-Balance Volume (OBV) tracks buying pressure to avoid "fakeouts." Liquidity Verification: Entries require a volume median ratio of 1.5x to ensure high-conviction momentum. Risk Management & Guardrails ZKSentinel prioritizes capital preservation through hard-coded "Sentinel Guardrails": Circuit Breaker: Automatic halt after 3 consecutive losses to prevent death spirals. Volatility Guard: Blocks new entries during extreme price swings exceeding 3%. Cooldown Logic: A mandatory 3-candle rest period after every exit to avoid choppy market noise. Precision Exit: Strict Stop-Loss (0.8%) and Take-Profit (1.5%) targets ensure a disciplined risk-to-reward ratio. ERC-8004 & Cryptographic Trust Every trade generates an ERC-8004 Validation Artifact. Using SHA-256 hashing, the agent creates a "fingerprint" of its strategy checkpoints (EMA deltas, volume ratios, and trade intent). This provides a transparent, tamper-proof audit trail for stakeholders without exposing proprietary weights. Roadmap Phase 1 (Current): Strategy hardening via CoinGecko API and ERC-8004 artifact logging. Phase 2 (Q4 2026): Transitioning artifacts into ZK-SNARKs to prove risk compliance while maintaining total strategy privacy and MEV resistance.