XSpoonAi AI DeFi Safety & Smart Execution Agent
Skill Overview
The AI DeFi Safety & Smart Execution Agent is an AI-powered Web3 skill that transforms a user’s DeFi intent (such as token swaps) into a safe, simulated, and explainable execution plan.
Instead of blindly executing transactions, the agent evaluates token risks, simulates execution paths, scores safety, and provides a clear recommendation before any on-chain action is taken.
Core Capabilities
1. DeFi Intent Understanding
The skill accepts high-level user intent such as:
- Swap token X to token Y
- Specify chain and amount
- Optional slippage and risk tolerance
This allows non-technical users or agents to express actions in a simple, human-readable format.
2. Token Safety & Risk Analysis
For the involved tokens, the agent analyzes:
- Liquidity depth
- Holder concentration
- Honeypot / blacklist behavior (heuristic or API-based)
- Centralization risks (admin privileges, pause/mint indicators)
The result is a structured list of risk flags and an intermediate token risk assessment.
3. Swap Simulation & Execution Feasibility
Before recommending execution, the skill:
- Simulates the swap route (DEX aggregation or heuristic simulation)
- Estimates expected output
- Evaluates slippage impact
- Estimates gas cost
- Flags potential MEV or instability risks
This ensures the action is feasible and transparent before execution.
4. Safety Scoring & Verdict Engine
All signals are aggregated into a single Safety Score (0–100) with a clear verdict:
- SAFE (80–100)
- CAUTION (50–79)
- DANGEROUS (<50)
This allows both humans and agents to make quick, informed decisions.
5. AI Explainability Layer
An AI reasoning layer converts technical risk signals into plain-English explanations, answering:
- Why the action is safe or risky
- What the primary dangers are
- What the user should be careful about
This makes the skill accessible even to non-expert users.
Input Schema (Example)
{
"action": "swap",
"from_token": "USDC",
"to_token": "ETH",
"amount": 500,
"chain": "Arbitrum",
"max_slippage": 1.0,
"risk_tolerance": "medium"
}