github-copilot
npx machina-cli add skill G1Joshi/Agent-Skills/github-copilot --openclawGitHub Copilot
GitHub Copilot is the enterprise standard. 2025 features Copilot Workspace (Idea-to-PR workflow) and Copilot Edits (multi-file awareness).
When to Use
- Enterprise: SOC2 compliant, IP indemnity.
- In-IDE: Works in VS Code, Visual Studio, JetBrains, Vim.
- CLI:
gh copilot suggest "how to tar a folder".
Core Concepts
Copilot Workspace
A dev environment where you describe a task, and Copilot generates a plan + code + tests, then opens a PR.
Copilot Chat
Chat interface aware of your open files and selection.
Ghost Text
The gray text completions as you type.
Best Practices (2025)
Do:
- Use Workspace for Issues: Start a task from a GitHub Issue -> "Open in Workspace".
- Comment Driven Dev: Write a comment
// Calculate fibonacci, wait for completion. - Context: Keep relevant files open. Copilot looks at open tabs for context.
Don't:
- Don't blindly accept: Verify security and logic.
References
Source
git clone https://github.com/G1Joshi/Agent-Skills/blob/main/skills/ai-ml/github-copilot/SKILL.mdView on GitHub Overview
GitHub Copilot is the enterprise standard for AI-assisted coding, featuring Copilot Workspace to generate task plans, code, and tests, and Copilot Edits for multi-file awareness. It also provides an in-editor Copilot Chat and Ghost Text for context-aware completions. This accelerates turning descriptions into working software while keeping security and reviews in mind.
How This Skill Works
It works by analyzing the code and open files in your IDE to surface contextual completions (Ghost Text). You can start a Copilot Workspace from an issue or task description; Copilot generates a plan, code, tests, and then opens a PR. Copilot Chat offers a context-aware chat that references your current files and selections, while the CLI lets you request quick suggestions with gh copilot.
When to Use It
- Enterprise-grade SOC2 compliance and IP indemnity
- In-IDE development across VS Code, Visual Studio, JetBrains, Vim
- Start a task from a GitHub Issue using Open in Workspace to generate a plan, code, tests, and a PR
- Use Copilot Edits for multi-file-aware changes
- Use the CLI for quick prompts like gh copilot suggest \"how to tar a folder\"
Quick Start
- Step 1: Install and enable GitHub Copilot in your IDE (VS Code, Visual Studio, JetBrains, Vim) and optionally via the CLI.
- Step 2: Open a GitHub Issue and select Open in Workspace to spawn a workspace task that will generate plan, code, tests, and a PR.
- Step 3: Use comments like // Calculate fibonacci or rely on Ghost Text to accept, modify, and commit the generated results.
Best Practices
- Use Workspace to kick off Issues from a GitHub Issue by selecting Open in Workspace and letting Copilot generate a plan, code, tests, and a PR
- Drive work with a comment like // Calculate fibonacci to trigger a completion
- Keep relevant files open so Copilot can see context from open tabs
- Verify security and logic; don't blindly accept generated code
- Review and test generated code and tests before merging
Example Use Cases
- Start a feature task from an Issue and let Copilot generate the plan, code, tests, and a PR via Workspace
- Type a comment // Calculate fibonacci to trigger a complete function
- Use Ghost Text to fill in boilerplate while coding a new module
- Refactor across multiple files with Copilot Edits for coordinated changes
- Use gh copilot suggest \"how to tar a folder\" for quick CLI snippets
