Trading agents are very useful for developing and backtesting quality trading strategies for actions taking in the real world. However, the existing trading agent research mainly focuses on simulation using artificial data and market models. As a result, the actionable capability of developed trading strategies is often limited. In this paper, we analyze such constraints on developing actionable trading strategies for trading agents. These points are deployed into developing a series of trading strategies for trading agents through optimizing, and enhancing actionable trading strategies. We demonstrate working case studies in large-scale of market data. These approaches and their performance are evaluated from both technical and business perspectives.