Data Inputs and Decision Variables

Data Sources Referenced by Each AI Agent

Below is a comprehensive table listing various data sources the AI agents may reference, along with which agents rely on them. Each agent specializes in particular aspects of the ecosystem and uses these inputs to make informed parameter adjustment proposals.

Data Source	Market Analysis (A1)	Treasury Mgmt (A2)	Demand & Inflation (A3)	Risk Mgmt (A4)	User Flow (A5)
Market Price (MID/USD)	X	X	X	X	X
Trading Volume & Liquidity (DEX data)	X		X	X	X
Volatility Metrics (Historical & Implied)	X		X	X
Order Book Depth (if available via Oracles/APIs)	X			X
Treasury Asset Balances (DAI, USDC, sUSDe, USDY)		X		X
MID Supply & Circulating Amounts	X	X	X	X	X
Bonding History (Bond Prices, Premium Accumulations)	X	X	X	X
Staking Participation Rates (Stakers, Unstakers)			X		X
APY History (Staking Rewards Changes Over Time)	X	X	X		X
Competing Protocol APYs & External Yield Benchmarks			X
Macro Indicators (BTC/ETH trends, stablecoin yields)	X	X	X	X
On-Chain User Behavior (Transaction Frequency, Wallet Age)			X		X
New User Registrations, Retention Rates					X
User Engagement Metrics (Time Staked, Redeem Patterns)			X		X
Oracle Feeds (Price Oracles, Peg Stability Signals)	X	X		X
Security Alerts & Suspicious Activity Logs				X
Historical Premium/Discount Data for Bonding	X	X	X	X
Sentiment or Off-Chain Community Indicators (if used)	X		X		X

Legend:

• X indicates the agent actively references and utilizes that data source.

• Blank cells mean the agent generally does not rely on that particular data source.

This table shows that:

• Market Analysis Agent (A1) focuses heavily on direct market conditions (price, volume, volatility, order book) and macro indicators.

• Treasury Management Agent (A2) concentrates on treasury asset balances, MID supply, and bond premium accumulations.

• Demand & Inflation Control Agent (A3) emphasizes user behavior related to staking, APY comparisons, and some market/macro conditions.

• Risk Management Agent (A4) keeps an eye on price and volatility data, oracle reliability, treasury status for risk buffers, and security alerts.

• User Flow Agent (A5) relies on user growth, retention, engagement metrics, and some market signals to recommend incentive adjustments (APY tweaks or bond discounts).

By collectively referencing these diverse data sources, the AI agents can form robust, multi-faceted proposals that the orchestration layer then merges into balanced and effective parameter adjustments.