So I Coded with Windsurf for a Week

In the video, the creator shares their week-long experience using Windsurf, an AI-powered coding assistant, while working on a Next.js frontend and Go Fiber backend project, completing 83 tasks with a mix of TypeScript and Markdown. They discuss both the benefits and limitations of Windsurf, including its unique Cascade feature for context-aware assistance, while also noting some frustrations with responsiveness and outdated training data.

In the video, the creator shares their experience using Windsurf, an AI-powered coding assistant, for a week while working on a Next.js frontend and Go Fiber backend project. Over six days, they completed 83 tasks, utilizing 19 user prompt credits and 50 flow action credits. The creator highlights that about 45% of their work involved TypeScript, along with some Markdown for README files. They plan to demonstrate live coding with Windsurf later in the video and mention their ongoing transition to using Neovim on NixOS, although they have not fully set it up yet due to time constraints.

The creator discusses their initial impressions of Windsurf, noting that it feels similar to Visual Studio Code (VS Code), which they have used for years. They mention that Windsurf appears to be a fork of VS Code, integrating an AI agent for enhanced coding assistance. They compare it to other AI coding tools like GitHub Copilot and Cursor, emphasizing Windsurf’s unique feature called Cascade, which allows users to initiate new projects or make changes across their entire codebase. This feature enables the AI to analyze and reference multiple files, providing context-aware assistance.

Throughout the week, the creator encountered both positive experiences and some frustrations with Windsurf. They appreciated the AI’s ability to analyze code and suggest modifications, but they also faced issues with the tool’s responsiveness and occasional bugs. For instance, they experienced difficulties with the accept/reject functionality when modifying code suggestions and had to restart the application multiple times due to glitches. Despite these annoyances, they found that Windsurf could effectively assist with coding tasks, especially when troubleshooting errors.

The creator also pointed out some limitations of Windsurf, such as outdated training data that led to incorrect version information for Next.js. They noted that while the AI can provide helpful suggestions, it may not always be up-to-date with the latest developments in programming languages and frameworks. However, they found the tool useful for running terminal commands and providing context-aware assistance, which enhanced their coding efficiency.

Looking ahead, the creator plans to continue using Windsurf alongside their Rust development projects and further explore Neovim. They express interest in refining their coding skills and adapting to new tools, promising to provide updates on their experience with Windsurf in the future. The video concludes with an invitation for viewers to subscribe for more content related to coding, development, and tool reviews.