GitHub Copilot Workspace lets you describe a task in natural language and get a fully planned, multi-file implementation. This guide covers how Workspace reasons about your repository and how to guide it effectively.
Copilot Workspace takes a GitHub issue or a task description and produces a full implementation plan: which files to create or modify, what changes to make, and why. You review and edit the plan before any code is written, which is fundamentally different from Copilot's inline completions.
Workspace understands your repository's existing patterns. It reads relevant files, follows your naming conventions, and writes code that fits your codebase rather than generic examples. The output quality is proportional to how well your existing code is structured.
Open any GitHub issue in your repository and click 'Open in Workspace'. Copilot reads the issue description, scans related files, and generates a plan. The plan lists specific files to create or edit with a description of what each change does.
Review the plan before proceeding. You can edit the plan — add steps, remove steps, or change the description of individual changes. The more specific you make the plan, the more accurate the implementation. Vague plans produce vague code.
When starting from a text prompt (not an issue), be specific about inputs, outputs, and constraints. Instead of 'Add search to the blog', write 'Add a full-text search endpoint GET /api/posts/search?q= that searches the title and body fields using a case-insensitive substring match. Return the same JSON structure as GET /api/posts.'
Reference existing files or functions when relevant: 'Follow the pattern in src/handlers/posts.ts when creating the search handler.' Workspace will read that file and mirror its structure.
After Workspace generates code, run it in the built-in terminal. If tests fail or the output is wrong, describe the issue in the Workspace chat: 'The search is case-sensitive. Fix it to be case-insensitive.' Workspace will revise the implementation.
Workspace sessions can be shared via URL — useful for code review or for handing off a task. All changes are on a branch; you create a pull request from the Workspace UI when satisfied.
Claude Code is Anthropic's agentic coding tool that runs in your terminal. It reads your entire codebase, writes and edits files, runs tests, and executes shell commands. This guide covers installation, configuration, and practical workflows.
Read guideCursor is a VS Code fork with deep AI integration. This guide covers installing Cursor, importing your existing VS Code settings, configuring AI models, and using Cursor's Composer for multi-file editing.
Read guide