My Experience With OpenCode: Powerful AI Coding Tool… Until It Broke in Large Repos

1. 1Build a small-scale project using OpenCode to understand its capabilities
2. 2Run the project with a limited scope to observe performance
3. 3Configure the project to increase scope and observe performance degradation
4. 4Test the limits of OpenCode with a large repository
5. 5Apply the lessons learned to adjust expectations and workflows

Software engineers and AI engineers can benefit from understanding the limitations of AI coding agents to effectively integrate them into their workflows and manage expectations

💡 The performance of AI coding agents is heavily dependent on the scope of the project, with larger scopes leading to significant performance degradation

🚨 AI coding agents like OpenCode can struggle with large scopes #AI #coding