Devin Desktop: AI-Powered Development (formerly Windsurf)

Build a small, practical app that uploads a CSV and displays a chart, while establishing clear specifications, tests, and a git-backed workflow for AI-powered development.

Develop a product requirements document for a windsurf web app that uploads a USD/EUR CSV, parses data, and renders an interactive time-series chart with tooltips, crosshair, legend, and axis labels.

Learn to build a to-do list from a software specification by breaking work into small features, creating a markdown todo file with acceptance criteria.

Implement the first feature by following the software specification and to-do list, complete phase one in one go, and launch a local server to preview the app in a browser.

Initialize the project for code versioning with git, streamline a project-specific gitignore, and manage ai-driven changes by accepting or rejecting file updates for reliability.

Rolling back changes to a working version demonstrates using git to discard buggy ai changes, inspect commit history, and restore a stable build.

Build phase two from the to-do list and software specification, implement the features, and verify with Windsurf preview that the chart initializes with sample data.

Automate testing with AI by selecting unit, integration, UI, and end-to-end tests, using Playwright for reliable end-to-end coverage. Avoid AI making code changes and manage test scope and artifacts.

Refactor a monolithic main.js into a modular React JS app with components, plan via a comprehensive to-do list, and implement end-to-end tests to prevent regressions during the AI-powered refactor.

Discover troubleshooting techniques for windsurf, including stopping stuck commands with the stop option or continue to resume by reissuing the command. Check the chat and intervene when no progress occurs.

Deploy your app with Windsurf Deploys in a one-click Netlify workflow, with Windsurf automatically generating the Netlify file for a preview deployment.

Navigate software development lifecycle with windsurf, using a software specification and to do list to guide AI, practice small increments with checks, emphasize testing, code reviews, refactoring, evolving plans.

Learn to control Windsurf by selecting appropriate models and writing clear rules with proper context. Use Windsurf preview or browser to reference elements and logs for more consistent code decisions.

Learn to select the right model for coding tasks by comparing standard GPT and reasoning models, evaluating credits and providers, and balancing agent and chat modes.

Learn how cascade rules in windsurf tailor AI behavior with workspace rules, global rules, and activation modes such as always on, manual, model decision, and glob (regex).

Provide context to speed up tasks by referencing files, directories, and specific lines of code. Use Cascade to explain code and troubleshoot in terminal with commands like npm run start.

Discover how to keep large language models current by using add web or pasting a URL, and leverage Windsurf docs to fetch up-to-date information for libraries and providers.

Explore Windsurf's IDE preview to start your app locally and view the live UI, then reference UI elements and rename a button to select file.

Discover the windsurf browser, a development-oriented browser with full dev tools, dom access, and logging. Integrated with Windsor, it supports element sending, screenshots, and seamless web development workflows.

Learn how to manage long conversations in cascade to preserve context and boost model performance by using memories, starting new chats, and referencing past conversations to keep your workspace organized.

Explore MCP workflows in Windsor, connect windsurf to external services via MCP, and explore the MCP store with ready-made integrations like GitHub.

Discover the model context protocol (MCP), an open standard that lets apps expose context and tools to LLMs via a plugin system, enabling access to external data sources and services.

Explore the MCP store, a marketplace for installing MCP servers, with two access methods, per-server variations, and steps to install, configure permissions, and manage tools.

Enable the GitHub MCP server to back up and push project code to a new GitHub repository using a personal access token. Install Docker locally and verify MCP integration.

See how to convert a to-do list into GitHub issues using the GitHub MCP tool, with default settings and repo data, creating phase-based issues with source and acceptance criteria.

Explore how team collaboration works in Windsurf, highlighting the main ideas. Learn about enterprise features that support teamwork and how to adopt Windsurf in your organization.

Windsurf offers team and enterprise plans with role-based access control, admin role assignment, and deployment controls; enterprise includes single sign-on with Google Workspace, Azure AD, or Okta.

Windsurf reviews connect to GitHub to automate code reviews, analyze changes up to 50 files, check for bugs, missing imports, conflicts, and missing tests, and comment with fixes.

Upload up to 50 documents to the knowledge base via Google Drive; cascade uses them as context for AI answers and code, including coding standards, API references, and project requirements.

Share complete chats within your team using conversation sharing, access prior discussions via add mention, and reference Cascade conversations to avoid duplicate work.

Deploy apps from Windsurf to a shared Netlify account, eliminating individual credentials and enabling Cascade to build and deploy under admin control.