> For the complete documentation index, see [llms.txt](https://aiw3.gitbook.io/aiw3/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://aiw3.gitbook.io/aiw3/trading-workflow.md).

# Trading Workflow

AIW3 structures onchain trading as a continuous, agent-coordinated workflow, replacing fragmented and manual processes with a unified execution loop.A standard workflow within AIW3 consists of:

#### 1. Market Observation

AIW3 continuously monitors onchain and cross-market data, including liquidity flows, whale activity, derivatives metrics, and market structure changes.

#### 2. Signal Generation

AI models transform raw data into structured signals, identifying trends, anomalies, and potential opportunities.

#### 3. agent-driven execution Construction

Based on these signals, strategies are generated or selected, with defined parameters such as entry/exit logic, leverage, and allocation rules.

#### 4. Risk Validation

Strategies are evaluated through backtesting, simulation, and real-time risk assessment, ensuring alignment with user-defined constraints.

#### 5. Execution & Capital Allocation

Validated strategies are deployed through non-custodial smart contracts.Capital is allocated via vaults, enabling automated and transparent execution.

#### 6. Monitoring & Optimization

Agents continuously monitor performance, adjust parameters, and optimize execution based on evolving market conditions.This workflow operates as a continuous service loop coordinated by agents, rather than a sequence of manual user actions.