CA2: Code-Aware Agent for Automated Game Testing

2026-05-13

AI Summary

This work addresses the limitations of existing automated game testing methods, which often struggle to achieve high code coverage, frequently overlook edge cases, and fail to leverage internal program information effectively. To overcome these challenges, the authors propose a novel reinforcement learning agent that uniquely utilizes the function call stack as a key guiding signal. By integrating game state and visual input into a multimodal environment representation, the agent is able to perform targeted exploration toward specific function paths. The approach combines dynamic code instrumentation with real-time call stack tracing, enabling more informed test generation. Experimental results across multiple game environments demonstrate that this method significantly outperforms baseline techniques that do not exploit code-level information, achieving higher test coverage and efficiency. These findings underscore the critical role of call stack signals in intelligent software testing.

Abstract

Automated game testing is important for verifying game functionality, but it remains a costly and time-consuming process. Manual testing often misses edge cases, and current automated methods struggle to provide full code coverage. Prior work has explored reinforcement learning (RL) for game testing, but without leveraging internal code signals such as the call stack. We present Code Aware Agent (CA2), which uses call stack information to learn effective testing strategies. The agent receives the current function call trace along with the game state and learns to reach specific target functions. We instrument two types of environments, 1) State-based and 2) Image-based, with support for efficient call stack extraction. Through experimental evaluation, we find that CA2 achieves consistent improvement over the non-code aware baselines, which does not leverage call stack information. Our results show that incorporating code signals like the call stack enables more effective and targeted game testing.

Problem

• automated game testing
• code coverage
• call stack
• edge cases
• software testing

Innovation

• code-aware testing
• call stack
• automated game testing
• reinforcement learning
• test coverage