Skip to main content
SURCHI’s intelligence layer is built around three specialized AI agents called Sentinels. Rather than a monolithic system, the Neural Sentinel Protocol distributes DeFi intelligence across three cooperating agents — each an expert in its domain, each operating within a defined permission boundary. Together, they form a fully autonomous DeFi intelligence layer purpose-built for Solana.

​
The Sentinel Triad

Three distinct agents, each with a specific role, designed to work in concert: Alpha Sentinel — Market Intelligence The researcher. Alpha Sentinel continuously scans on-chain data, DEX activity, whale wallet movements, and social sentiment to surface opportunity and risk signals before the market prices them in. Liquidity Sentinel — Liquidity Protection The risk manager. Liquidity Sentinel watches over liquidity pools, LP positions, protocol reserves, and slippage conditions — providing a real-time picture of liquidity health and alerting you before risk becomes loss. Execution Sentinel — Autonomous On-Chain Action The trader. Execution Sentinel translates validated intelligence into signed, on-chain Solana transactions — executing swaps, managing positions, and carrying out multi-step strategies within your defined risk limits.

Alpha Sentinel

Market intelligence, opportunity detection, and whale signal monitoring across the Solana ecosystem.

Liquidity Sentinel

Real-time liquidity monitoring, impermanent loss tracking, and automated protection for your positions.

Execution Sentinel

Autonomous on-chain execution with configurable risk controls, slippage management, and full auditability.

​
How Sentinels Cooperate

No single Sentinel acts independently on high-impact decisions. The three agents are designed to operate as a coordinated system with built-in checks and balances:
  1. Alpha Sentinel surfaces an opportunity — it identifies a signal meeting your configured confidence threshold and flags it for action.
  2. Liquidity Sentinel validates the risk — before any trade is queued, Liquidity Sentinel checks pool health, available depth, slippage conditions, and your current exposure.
  3. Execution Sentinel proceeds only when both conditions are met — if Alpha has a signal and Liquidity approves the conditions, Execution carries out the strategy within your permission boundaries.
This sequenced cooperation means no trade is executed on signal alone. Every execution requires a clean bill of health from the risk layer — creating a natural check-and-balance system that mirrors how professional trading desks operate. If Liquidity Sentinel flags a risk condition, execution is paused regardless of how strong the Alpha signal is. You stay in control, and the system enforces caution automatically.

​
Sentinel Permissions

Each Sentinel operates within a permission model you define. You determine what each agent is allowed to do — and what it is not.
  • Alpha Sentinel can be configured to alert only, or to automatically forward high-confidence signals to the Execution queue.
  • Liquidity Sentinel can be configured to alert only, or to take automated protective actions (such as withdrawing LP positions) when thresholds are breached.
  • Execution Sentinel requires explicit authorization for every asset and action type. It will never exceed your configured position size, daily cap, or asset whitelist.
You can adjust permissions at any time from the SURCHI dashboard. The Sentinels operate within whatever boundaries you set — expanding or restricting autonomy as your confidence grows.
Think of the three Sentinels as a trading desk: a researcher (Alpha), a risk manager (Liquidity), and a trader (Execution) — all operating autonomously within your guidelines.