GitHub Copilot

GitHub-native AI coding platform spanning IDE, CLI, repository instructions, MCP, plugins, skills, code review, and cloud-agent workflows.

Type: IDE assistant / CLI / PR tooling / cloud agent

Best for: GitHub-heavy teams that want one platform across editor, terminal, repository, review, and agent workflows

Official docs: https://docs.github.com/en/copilot

Last verified: 2026-08-09

Status: Documentation verified

Verification scope: Official GitHub Docs for Copilot CLI installation, command reference, plan mode, tool approvals, MCP, plugins, skills, custom instructions, and repository initialization were reviewed. Commands were not executed locally in this repository.

60-Second Cheat Sheet

Safe default: use plan mode for broad work, approve mutation tools narrowly, and prefer --allow-tool / --deny-tool over --allow-all-tools when automation needs extra permissions.

Five Prompts Worth Bookmarking

Understand a repository

Give me an overview of this repository: architecture, entry points, build/test commands, important GitHub workflows, and one safe first improvement. Do not edit files.

Plan a feature

Analyze the relevant code and tests, then create an implementation plan with files to change, validation commands, risks, and rollback points. Do not write code until the plan is approved.

Review a pull request

Review the current diff for correctness, regressions, security issues, and missing tests. List actionable findings by severity and cite the affected files.

Prepare repository instructions

Review this repository and propose concise Copilot instructions covering architecture, build/test/lint commands, coding conventions, risky directories, and validation expectations. Keep repository-specific facts only.

Debug CI

Inspect the failing workflow/check logs, identify the first root-cause failure rather than downstream noise, propose the smallest fix, and list the exact validation needed before changing anything.

What GitHub Copilot Is Best At

• organizations already centered on GitHub repositories and pull requests
• one workflow spanning editor, CLI, code review, and GitHub.com agent surfaces
• repository initialization through generated custom instructions
• plan-first terminal workflows
• MCP, plugins, skills, custom agents, hooks, and GitHub integrations
• teams that want organization policy to gate AI capabilities centrally

Installation

npm

npm install -g @github/copilot

GitHub currently documents Node.js 22 or later as the prerequisite for npm installation.

Windows

winget install GitHub.Copilot

macOS and Linux

brew install --cask copilot-cli

GitHub also documents additional install paths. Use the current GitHub Docs installation page as the source of truth for platform-specific updates.

First Session

cd /path/to/your/repository
copilot

Then authenticate if needed:

/login

Copilot asks you to confirm that you trust the current directory for AI use. During a session it may read, modify, and execute files in and below that directory, subject to approvals and configuration.

Start safely:

Give me an overview of this project and its build/test commands. Do not modify files.

Daily CLI Commands

copilot
copilot -p "summarize the current diff"
copilot --continue
copilot init
copilot mcp
copilot help permissions

Useful in-session commands include:

/plan
/model
/permissions
/init
/instructions
/plugins
/ide
/help

GitHub's current CLI also includes parallel subagent execution with /fleet, plugins, skills, custom agents, hooks, and remote/session capabilities. Use the CLI reference for the exact version installed in your environment.

Plan Mode

Copilot CLI has a dedicated plan mode for analyzing a task and building a structured implementation plan before writing code.

Use:

/plan Implement the new authentication flow with tests and migration notes

This is a useful default for:

• cross-cutting changes
• unfamiliar repositories
• migrations
• production-impacting changes
• work that needs agreement on scope before implementation

Repository Instructions

The CLI can initialize repository instructions with:

copilot init

or inside a session:

/init

The standard repository-wide file is:

.github/copilot-instructions.md

Path-specific instructions can live under:

.github/instructions/**/*.instructions.md

Current Copilot CLI also loads and merges several instruction formats, including AGENTS.md, CLAUDE.md, and GEMINI.md in supported locations.

For teams, decide which formats are canonical so the same rule is not duplicated across files.

Permission Model

Copilot asks before using tools that can modify or execute files unless permissions have already been granted for the session or configured through CLI flags/settings.

Important automation flags include:

• --allow-tool — pre-approve a specific tool or command pattern
• --deny-tool — block a tool; deny takes precedence
• --allow-all-tools — broadly removes per-tool approval prompts

Prefer narrow patterns such as a specific git or test command over broad shell access.

Do not use --allow-all-tools as a normal developer default: GitHub explicitly warns that this can give Copilot the same local file and shell access as the user account running it.

MCP, Plugins, and Skills

Copilot CLI now has first-class support for:

• MCP servers
• plugins and plugin marketplaces
• agent skills
• custom agents
• hooks

Persistent MCP configuration can be managed through the CLI, while /plugins provides a dashboard for plugins, MCP servers, and skills.

Treat every extension mechanism as a separate trust decision. A base Copilot approval does not automatically mean every MCP server, skill, plugin, or hook should be approved.

Team Adoption Checklist

• [ ] enable Copilot CLI through organization policy only after review
• [ ] standardize .github/copilot-instructions.md
• [ ] decide whether AGENTS.md/CLAUDE.md/GEMINI.md are also supported
• [ ] default complex work to plan mode
• [ ] define allowed/denied tools for automated usage
• [ ] review MCP servers, plugins, skills, and hooks separately
• [ ] keep CI and PR validation as the final enforcement layer

Security Notes

• trusting a directory makes its content available to the CLI workflow
• session approvals can widen command access significantly
• --allow-all-tools removes important checkpoints
• multiple instruction files are merged, so conflicting guidance can be difficult to notice
• MCP/plugins/skills/hooks expand the execution and data boundary
• organization policy should gate features that are not ready for broad use

Alternatives

• Cursor for a more editor-first multi-model environment
• OpenAI Codex for OpenAI-centered terminal, local review, and automation workflows
• Claude Code for terminal-first work with strong permission and project-memory patterns
• Gemini CLI for a Google-backed terminal workflow with sandbox and context controls

See the Comparison Matrix for a side-by-side shortlist.

Verification Status

• Status: Documentation verified
• Last verified: 2026-08-09
• Scope: CLI install, authentication, plan mode, custom instructions, approvals, MCP, plugins, skills, and command reference reviewed against current GitHub Docs
• Not locally tested: CLI installation, login, organization policy changes, editor behavior, plugin installation, and cloud-agent workflows

Sources

• https://docs.github.com/en/copilot
• https://docs.github.com/en/copilot/how-tos/copilot-cli/cli-getting-started
• https://docs.github.com/en/copilot/reference/copilot-cli-reference/cli-command-reference
• https://docs.github.com/en/copilot/concepts/agents/copilot-cli/about-copilot-cli
• https://docs.github.com/en/copilot/how-tos/copilot-cli/set-up-copilot-cli/configure-copilot-cli
• https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/overview
• https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/add-custom-instructions