Transaction Types

Nexgent AI tracks three types of transactions for each agent: DEPOSIT, SWAP, and BURN. Each transaction type serves a specific purpose and is recorded with detailed information for tracking and analytics.

Overview

All agent transactions are stored in the agent_transactions table and include metadata (ID, agent, wallet, timestamp), transaction type (DEPOSIT, SWAP, BURN), value information (USD value, token amounts, prices), and related data (signal ID, destination address).

Transactions update balances atomically and trigger cache updates for real-time dashboard display.

Transaction Types

DEPOSIT

Purpose: Add funds to an agent's wallet

Aspect	Description
When it occurs	User deposits funds (simulation or live)
Balance effect	Increases wallet balance
Token	Typically SOL (native token)
Execution	Simulation: Virtual balance update Live: On-chain balance sync

Fields: transactionType (DEPOSIT), transactionValueUsd, inputMint, inputAmount, destinationAddress

Detailed view of a deposit transaction showing confirmation status and amount

SWAP

Purpose: Exchange one token for another (purchase or sale)

Aspect	Description
When it occurs	Agent executes a token swap (purchase or sale)
Balance effect	Decreases input token, increases output token
Execution	Simulation: Mock swap result Live: On-chain swap via Jupiter Aggregator
Details	Includes input/output amounts, prices, slippage, fees

Fields:

transactionType: "SWAP"
transactionValueUsd: Swap value in USD
inputMint: Token being sold (mint address)
inputSymbol: Token symbol being sold
inputAmount: Amount of input token
inputPrice: Price per unit of input token
outputMint: Token being bought (mint address)
outputSymbol: Token symbol being bought
outputAmount: Amount of output token received
outputPrice: Price per unit of output token
slippage: Slippage percentage
priceImpact: Price impact percentage
fees: Transaction fees (in USD)
isDca: Whether this is a DCA purchase
signalId: Signal that triggered the swap (if applicable)
transactionHash: On-chain transaction hash (live mode only)

Purchase swap (SOL → Token): inputMint = SOL, outputMint = Token
Sale swap (Token → SOL): inputMint = Token, outputMint = SOL

Detailed view of a swap transaction showing route, slippage, and price impact

BURN (simulation only)

Purpose: Destroy tokens (remove from circulation)

Aspect	Description
When it occurs	Tokens are burned/destroyed
Balance effect	Decreases token balance
Execution	On-chain burn transaction (live mode)
Use case	Token destruction, deflationary mechanisms

Fields: transactionType (BURN), transactionValueUsd, inputMint, inputAmount, inputPrice

Transaction Structure

All transactions share common fields:

Field	Type	Description
id	string (UUID)	Unique transaction identifier
agentId	string (UUID)	Agent that owns the transaction
walletAddress	string	Wallet address (simulation or live)
transactionType	enum	DEPOSIT, SWAP, or BURN
transactionValueUsd	decimal	Transaction value in USD
transactionTime	datetime	When transaction occurred
destinationAddress	string (optional)	Destination wallet address
signalId	number (optional)	Signal that triggered transaction
createdAt	datetime	When record was created
updatedAt	datetime	Last update timestamp

SWAP-specific fields:

Input/output token details (mint, symbol, amount, price)
Slippage and price impact
Fees
Transaction hash (live mode)
DCA flag

Transaction Lifecycle

Creation

Transactions are created when:

User action - Deposit via dashboard
Agent trading - SWAP executed by agent
External event - On-chain transaction detected (live deposits)

Balance Update

Transactions update balances atomically (balance update and transaction creation happen together). After commit: Redis cache updated, dashboard displays real-time balance, WebSocket notifies clients.

Guarantees: Atomicity, consistency (balances match transaction history), isolation (concurrent transactions don't interfere)

Transaction Viewing

View transactions in the dashboard:

Filtering: By type (DEPOSIT, SWAP, BURN), token, signal, date range, wallet (simulation or live)

Transaction details: Full transaction information, related signal (if applicable), input/output token details (SWAP), slippage/fees/price impact (SWAP), on-chain transaction hash (live SWAP)

Common Patterns

Purchase Flow

Trigger: Trading signal received

Steps:

Agent evaluates signal (eligibility check)
SWAP transaction created: SOL → Token
Balances updated (SOL decreases, Token increases)
Position created (if new token)
Transaction recorded with signal ID

Transaction details:

Type: SWAP
Input: SOL amount
Output: Token amount
signalId: Signal that triggered purchase
isDca: true if DCA purchase, false otherwise

Transaction Relationships

Signals

SWAP transactions link to signals via signalId: purchase SWAPs use the signal that triggered the trade, sale SWAPs use the original purchase signal from the position. DEPOSIT and BURN are not signal-related.

Positions

Purchase SWAP: Creates new position (if token not held) or updates existing position (DCA)
Sale SWAP: Closes position and creates historical swap record

Historical Swaps

Sale SWAP transactions create historical swap records linking purchase and sale transactions (trade pairs).

Best Practices

Monitor transaction volume: Track frequency per type, identify unusual patterns, monitor balance changes.

Review transaction details: Check slippage and fees (SWAP), verify on-chain hashes (live SWAP), confirm signal associations.

Balance reconciliation: Verify transaction history matches balances, reconcile with on-chain data (live mode).

Transaction analysis: Use for performance analysis, fee tracking, strategy optimization, tax reporting.

Signal Processing Flow Historical Swaps