{"id":48842956,"url":"https://github.com/nishanil/copilot-guide","last_synced_at":"2026-04-15T03:04:04.617Z","repository":{"id":339545453,"uuid":"1162373846","full_name":"nishanil/copilot-guide","owner":"nishanil","description":"A simple guide to every GitHub Copilot customization and agentic feature — instructions, agents, skills, MCP, coding agent, and more.","archived":false,"fork":false,"pushed_at":"2026-03-08T23:20:51.000Z","size":276,"stargazers_count":2,"open_issues_count":0,"forks_count":1,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-03-09T00:56:41.918Z","etag":null,"topics":["copilot","github"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/nishanil.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-20T07:18:52.000Z","updated_at":"2026-03-08T23:20:55.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/nishanil/copilot-guide","commit_stats":null,"previous_names":["nishanil/copilot-guide"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/nishanil/copilot-guide","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nishanil%2Fcopilot-guide","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nishanil%2Fcopilot-guide/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nishanil%2Fcopilot-guide/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nishanil%2Fcopilot-guide/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/nishanil","download_url":"https://codeload.github.com/nishanil/copilot-guide/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nishanil%2Fcopilot-guide/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31824118,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-14T18:05:02.291Z","status":"online","status_checked_at":"2026-04-15T02:00:06.175Z","response_time":63,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["copilot","github"],"created_at":"2026-04-15T03:04:03.949Z","updated_at":"2026-04-15T03:04:04.592Z","avatar_url":"https://github.com/nishanil.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cdiv align=\"center\"\u003e\n\n# 🤖 Understanding GitHub Copilot\n\n**A comprehensive guide to every GitHub Copilot customization and agentic feature.**\n\nWhat each does · When you'd use it · How they fit together\n\n[![GitHub Copilot](https://img.shields.io/badge/GitHub-Copilot-blue?logo=github)](https://github.com/features/copilot)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)\n[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](https://github.com/nishanil/copilot-guide/pulls)\n\n\u003c/div\u003e\n\n\u003cp class=\"gh-pages-link\"\u003e📖 \u003cstrong\u003e\u003ca href=\"https://nishanil.github.io/copilot-guide/\"\u003eRead this guide on GitHub Pages\u003c/a\u003e\u003c/strong\u003e for a better reading experience.\u003c/p\u003e\n\n## Who is this guide for?\n\nThis is a single-page reference to get you onboarded quickly with GitHub Copilot's customization and agentic features. It covers what each feature does, when to use it, and how they fit together — so you can start building without reading dozens of docs pages first.\n\nFor detailed documentation, see the [Further Reading](#further-reading) section at the end.\n\n\u003e ⚠️ **Note:** This guide was written with the help of an AI agent and is kept up to date by an [agentic workflow](.github/workflows/check-copilot-updates.md) that checks for new Copilot features daily. I've done my best to keep it accurate, but always double-check details against the [official documentation](https://docs.github.com/en/copilot). If something looks off, it's probably the agent's fault 🤖\n\n## At a Glance\n\n\u003cdiv class=\"ascii-glance\" markdown\u003e\n\n```\n┌─────────────────────────────────────────────────────────────────────┐\n│                                                                     │\n│  📄 CUSTOMIZATION — files that shape how Copilot behaves            │\n│                                                                     │\n│  Instructions        Always-on project context \u0026 conventions        │\n│  Path-specific       Different rules for different parts            │\n│  Prompt files        Reusable recipes for repeated tasks            │\n│  Skills              Packaged expertise, auto-loaded on demand      │\n│  MCP Servers         Connect to external tools, databases, APIs     │\n│  Hooks               Automate lifecycle events (format, lint, startup) │\n│  Custom Agents       Named specialists with personas                │\n│  Plugins             Bundle all of the above into one package       │\n│                                                                     │\n├─────────────────────────────────────────────────────────────────────┤\n│                                                                     │\n│  ⚡ AGENTIC — runtime capabilities for delegation \u0026 parallelism     │\n│                                                                     │\n│  Running Parallel    Auto-delegates work; /fleet for explicit        │\n│    Tasks \u0026 /fleet    parallel execution via subagents                │\n│  Autopilot Mode      Hands-off autonomous task completion (CLI)     │\n│  Coding Agent        Autonomous cloud agent — issues in, PRs out    │\n│  Agentic Workflows   Automate repo tasks via GitHub Actions + AI    │\n│  Mission Control     Dashboard to manage coding agents at scale     │\n│  Agents Tab          Repo-level UI for agent tasks and tracking     │\n│                                                                     │\n├─────────────────────────────────────────────────────────────────────┤\n│                                                                     │\n│  🧠 PLATFORM — built-in intelligence \u0026 extensibility                │\n│                                                                     │\n│  Agentic Memory      Per-repo persistent memory across sessions     │\n│  Copilot Spaces      Project-specific knowledge containers          │\n│  Copilot SDK         Build your own agents as code (Node/Python/Go) │\n│                                                                     │\n├─────────────────────────────────────────────────────────────────────┤\n│                                                                     │\n│  🌐 COMMUNITY — open-source projects built on Copilot              │\n│                                                                     │\n│  Squad               Persistent AI team with memory \u0026 identity      │\n│  awesome-copilot     Curated collection of patterns \u0026 resources     │\n│                                                                     │\n└─────────────────────────────────────────────────────────────────────┘\n```\n\n\u003c/div\u003e\n\n\u003cdiv class=\"svg-glance\"\u003e\n\u003c!-- SVG generated at build time for GitHub Pages; not visible on GitHub.com --\u003e\n\u003c/div\u003e\n\n\u003cdiv class=\"readme-toc\" markdown\u003e\n\n---\n\n## Table of Contents\n\n- [Customization Features](#customization-features)\n  - [Custom Instructions](#custom-instructions)\n  - [Path-Specific Instructions](#path-specific-instructions)\n  - [Prompt Files](#prompt-files)\n  - [Skills](#skills)\n  - [MCP Servers](#mcp-servers)\n  - [Hooks](#hooks)\n  - [Custom Agents](#custom-agents)\n  - [Plugins](#plugins)\n- [Agentic Features](#agentic-features)\n  - [Running Parallel Tasks](#running-parallel-tasks)\n  - [Autopilot Mode](#autopilot-mode)\n  - [Copilot Coding Agent](#copilot-coding-agent)\n  - [Agentic Workflows](#agentic-workflows)\n  - [Mission Control](#mission-control)\n  - [Agents Tab](#agents-tab)\n- [Platform Features](#platform-features)\n  - [Agentic Memory](#agentic-memory)\n  - [Copilot Spaces](#copilot-spaces)\n  - [Copilot SDK](#copilot-sdk)\n- [Community Projects](#community-projects)\n  - [Squad](#squad)\n  - [awesome-copilot](#awesome-copilot)\n- [Best Practices](#best-practices)\n- [Quick Reference](#quick-reference)\n- [Further Reading](#further-reading)\n\n---\n\n\u003c/div\u003e\n\n## Customization Features\n\n\u003e Files you add to your repo (or user config) that shape how Copilot behaves.\n\n### Custom Instructions\n\nAlways-on project context. Copilot reads this on **every** interaction — chat, completions, code review.\n\n\u003e **When you need it:** Copilot doesn't know your stack, conventions, or preferences.\n\n**📁 Location:** `.github/copilot-instructions.md`\n\n```markdown\n# Project: RecipeShare\n- Framework: React 19 + Hono backend\n- Styling: Tailwind CSS (no CSS modules)\n- Language: TypeScript strict mode\n- Use `const` over `let`, never `var`\n- Tests: Vitest + React Testing Library\n- Database: Drizzle ORM with PostgreSQL\n```\n\n**Effect:** Every Copilot response now respects these rules. No agent needed. No tools. Just context.\n\n| | |\n|---|---|\n| **Scope** | Repository-wide, always-on |\n| **Applies to** | Chat, completions, code review — everything |\n| **Commit it** | Yes — the whole team benefits |\n\n---\n\n### Path-Specific Instructions\n\nOverride or refine rules for specific files or folders.\n\n\u003e **When you need it:** Your backend code needs different rules than your frontend. Test files shouldn't be as strict.\n\n**📁 Location:** `.github/instructions/*.instructions.md`\n\n```\n.github/\n├── copilot-instructions.md              # global\n├── instructions/\n│   ├── frontend.instructions.md         # applyTo: \"src/client/**\"\n│   ├── backend.instructions.md          # applyTo: \"src/server/**\"\n│   └── tests.instructions.md            # applyTo: \"**/*.test.ts\"\n```\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003ebackend.instructions.md\u003c/code\u003e\u003c/summary\u003e\n\n```markdown\n---\napplyTo: \"src/server/**\"\n---\n- All API routes return JSON with { data, error } envelope\n- Use Zod for request validation\n- Every endpoint needs rate limiting middleware\n```\n\n\u003c/details\u003e\n\n**Effect:** Backend files get global + backend instructions. Frontend files get global + frontend instructions. They don't bleed into each other.\n\n| | |\n|---|---|\n| **Scope** | Targeted via `applyTo` glob patterns |\n| **Merging** | Path-specific merges with global (doesn't replace) |\n\n---\n\n### Prompt Files\n\nReusable task templates you invoke explicitly via `/` commands in chat.\n\n\u003e **When you need it:** You keep doing the same multi-step tasks — \"create a new API endpoint\" always involves route + validation + test + migration.\n\n**📁 Location:** `.github/prompts/*.prompt.md`\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003enew-endpoint.prompt.md\u003c/code\u003e\u003c/summary\u003e\n\n```markdown\nCreate a new Hono API endpoint:\n1. Add route in src/server/routes/\n2. Add Zod schema in src/server/schemas/\n3. Add Drizzle migration in drizzle/migrations/\n4. Add Vitest integration test\n5. Update the OpenAPI spec\n```\n\n\u003c/details\u003e\n\n**Usage:** Type `/new-endpoint` in Copilot chat → it runs the full recipe.\n\n| | |\n|---|---|\n| **Scope** | On-demand — invoked explicitly, never auto-loaded |\n| **Difference from instructions** | Prompts are manual recipes; instructions are always-on context |\n\n---\n\n### Skills\n\nModular knowledge packages with a `SKILL.md` that Copilot loads on-demand when relevant.\n\n\u003e **When you need it:** Your deployment or migration process is complex and you want Copilot (including the coding agent) to know how to do it without you explaining every time.\n\n**📁 Location:** `.github/skills/{skill-name}/SKILL.md`\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003edrizzle-migrations/SKILL.md\u003c/code\u003e\u003c/summary\u003e\n\n```markdown\n---\nname: drizzle-migrations\ndescription: Create, review, and apply Drizzle ORM database migrations.\n---\n\n# Drizzle Migration Skill\n\n## When to use\nWhen creating or modifying database tables.\n\n## Steps\n1. Run `npx drizzle-kit generate` to create migration\n2. Review generated SQL in drizzle/migrations/\n3. Run `npx drizzle-kit migrate` to apply\n4. Update src/server/db/schema.ts with new types\n\n## Gotchas\n- Always check for existing data before DROP COLUMN\n- Use `default()` for new non-nullable columns on existing tables\n```\n\n\u003c/details\u003e\n\n**Effect:** When Copilot detects a task related to migrations, it loads this skill automatically. Works across CLI, VS Code, and the coding agent.\n\n| | |\n|---|---|\n| **Scope** | Auto-loaded when the task domain matches |\n| **Difference from prompts** | Skills are auto-detected; prompts are manually invoked |\n\n---\n\n### MCP Servers\n\n[Model Context Protocol](https://docs.github.com/en/copilot/how-tos/provide-context/use-mcp/extend-copilot-chat-with-mcp) servers connect Copilot to external tools — databases, APIs, cloud services, issue trackers — so agents can read and write to systems beyond your local files.\n\n\u003e **When you need it:** Your app needs to interact with other systems, ex: query PostgreSQL, check GitHub issues, or interact with cloud storage. Without MCP, you'd have to copy-paste data into chat.\n\n**📁 Location:** `.vscode/mcp.json` (VS Code) or `~/.copilot/mcp-config.json` (CLI)\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample configuration\u003c/summary\u003e\n\n```jsonc\n{\n  \"servers\": {\n    \"github\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@modelcontextprotocol/server-github\"],\n      \"env\": { \"GITHUB_TOKEN\": \"${GITHUB_TOKEN}\" }\n    },\n    \"postgres\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@modelcontextprotocol/server-postgres\"],\n      \"env\": { \"DATABASE_URL\": \"${DATABASE_URL}\" }\n    },\n    \"filesystem\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@modelcontextprotocol/server-filesystem\", \"./docs\"]\n    }\n  }\n}\n```\n\n\u003c/details\u003e\n\n| Without MCP | With MCP |\n|---|---|\n| \"Here's my schema, can you write a query?\" | Agent reads the schema directly from PostgreSQL |\n| \"The issue says...\" (you paste it) | Agent reads the issue via GitHub MCP server |\n| \"Upload to cloud storage\" (you do it manually) | Agent calls cloud storage API through an MCP server |\n\n| | |\n|---|---|\n| **Scope** | Available to all agents in the session |\n| **Discovery** | VS Code has a built-in MCP server gallery (search `@mcp` in Extensions) |\n| **Security** | Servers run locally — your credentials stay on your machine |\n| **OAuth / API keys** | MCP servers can request you to visit a URL for out-of-band auth flows (e.g. OAuth, API key entry) |\n\n---\n\n### Hooks\n\nCustom scripts that run automatically at specific lifecycle events — like pre-commit formatting, post-generation linting, or startup prompts.\n\n\u003e **When you need it:** You want every code generation to be auto-formatted with Prettier, or every commit to run lint checks, without remembering to do it manually.\n\n**📁 Location:** `.github/hooks/` (repo-level) or `~/.copilot/hooks/` (personal, user-level)\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — repo-level \u003ccode\u003e.github/hooks/hooks.json\u003c/code\u003e\u003c/summary\u003e\n\n```jsonc\n{\n  \"hooks\": {\n    \"post-edit\": {\n      \"command\": \"npx prettier --write ${file}\"\n    },\n    \"pre-commit\": {\n      \"command\": \"npm run lint:fix\",\n      \"timeout\": 30\n    },\n    \"startup\": {\n      \"prompt\": \"/compact Summarize recent changes in RecipeShare\"\n    }\n  }\n}\n```\n\n\u003c/details\u003e\n\n**Hook types:**\n\n| Event | Trigger |\n|---|---|\n| `post-edit` | After Copilot edits a file |\n| `pre-commit` | Before a git commit |\n| `startup` | When a CLI session starts — auto-submits a prompt or slash command |\n\n**Config notes:**\n\n- Use `\"command\"` as a **cross-platform alias** for `bash`/`powershell` shell commands — works on all platforms without separate entries\n- `\"timeout\"` is accepted as an alias for `\"timeoutSec\"` for readable config\n- Personal hooks (`~/.copilot/hooks/`) apply across all repos; repo-level hooks (`.github/hooks/`) are scoped to that repo\n\n| | |\n|---|---|\n| **Scope** | Runs automatically at lifecycle events — no manual invocation |\n| **Personal hooks** | `~/.copilot/hooks/` — applies to all repos on your machine |\n| **Difference from skills** | Skills are knowledge Copilot reads; hooks are scripts Copilot runs |\n\n---\n\n### Custom Agents\n\nA named agent with a persona, specific tools, and specialized behavior.\n\n\u003e **When you need it:** You want a dedicated \"security reviewer\" that always checks for auth issues, SQL injection, etc. — not just general Copilot.\n\n**📁 Location:** `.github/agents/{name}.agent.md`\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003esecurity-reviewer.agent.md\u003c/code\u003e\u003c/summary\u003e\n\n```markdown\n---\nname: Security Reviewer\ndescription: Reviews code for security vulnerabilities\n---\nYou are a security-focused code reviewer for RecipeShare.\n\nAlways check for:\n- SQL injection (even with Drizzle ORM — check raw queries)\n- XSS in React components (dangerouslySetInnerHTML)\n- Auth bypass in API routes (missing middleware)\n- File upload validation (type, size, path traversal)\n- Rate limiting on auth endpoints\n- Secrets in code or config\n\nOutput format: list findings as HIGH/MEDIUM/LOW with file:line references.\n```\n\n\u003c/details\u003e\n\n**Usage:** Select \"Security Reviewer\" from `/agents` in Copilot. It's a separate selectable agent, not always-on.\n\n| | |\n|---|---|\n| **Scope** | Explicitly selected by the user — not auto-loaded |\n| **Difference from instructions** | Agents are interactive personas; instructions are passive context |\n\n---\n\n### Plugins\n\nInstallable packages that bundle agents, skills, hooks, and MCP server configs into a single distributable unit. ([Docs](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/plugins-creating))\n\n\u003e **When you need it:** Your team has built a great set of agents, skills, hooks, and MCP configs. Instead of copying files between repos, you package them as a plugin that anyone can install.\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample structure\u003c/summary\u003e\n\n```\nmy-copilot-plugin/\n├── plugin.json              # Manifest — what's in the package\n├── agents/\n│   └── security-reviewer.agent.md\n├── skills/\n│   └── drizzle-migrations/\n│       └── SKILL.md\n├── hooks/\n│   └── post-edit-prettier.json\n└── mcp/\n    └── postgres-config.json\n```\n\n\u003c/details\u003e\n\n```bash\n# Install from a GitHub repo\ncopilot plugin install github:myorg/recipeshare-toolkit\n\n# List installed plugins\ncopilot plugin list\n```\n\nEnable plugins automatically at startup by listing them in your CLI config:\n\n```jsonc\n// ~/.copilot/settings.json\n{\n  \"enabledPlugins\": [\n    \"github:myorg/recipeshare-toolkit\"\n  ]\n}\n```\n\n| | |\n|---|---|\n| **Location** | Installed into `~/.copilot/plugins/` or project-local |\n| **Scope** | All bundled agents/skills/hooks/MCP become available |\n| **Auto-install** | List in `enabledPlugins` in `~/.copilot/settings.json` to install automatically at startup |\n| **Difference from skills** | A skill is one knowledge file; a plugin is a complete toolkit |\n\n---\n\n## Agentic Features\n\n\u003e Runtime capabilities — how Copilot delegates work, runs in parallel, and operates autonomously.\n\n### Running Parallel Tasks\n\nWhen your request involves multiple independent jobs, the Copilot agent automatically breaks it into smaller pieces and delegates them to **child agents** (subagents) running in parallel. This works in both **Copilot CLI** and **VS Code**. In the CLI you can also use the **`/fleet`** slash command to explicitly request parallel execution.\n\n\u003e **When you need it:** Your request naturally involves multiple independent jobs — running tests while linting, scanning different parts of the codebase, or checking several things at once.\n\n#### What you say → What happens\n\n```\nYou:  \"Run the tests, lint the code, and check for security issues in RecipeShare.\"\n```\n\nBehind the scenes the agent spawns three child agents that work **simultaneously**:\n\n```\n┌─ child 1 ─────────────────┐  ┌─ child 2 ──────────────┐  ┌─ child 3 ──────────────────┐\n│ npm test                   │  │ npm run lint            │  │ Scan routes for missing    │\n│ (runs full test suite)     │  │ (checks style issues)   │  │ auth middleware             │\n└────────────────────────────┘  └─────────────────────────┘  └────────────────────────────┘\n```\n\nYou see progress updates as each child finishes, then a consolidated summary — no extra steps on your part.\n\n#### The `/fleet` slash command (CLI)\n\nUse `/fleet` in the CLI to explicitly tell Copilot to break a task into subtasks and run them in parallel via subagents. The main agent acts as an **orchestrator** — it analyses the prompt, maps dependencies, and fans out independent work.\n\n```\n\u003e /fleet Implement the recipe rating feature: add the database migration,\n  create the API endpoint, write unit tests, and update the OpenAPI spec.\n```\n\nEach subagent gets its **own context window**, so individual subtask context doesn't crowd the main session.\n\n**Model selection per subagent** — By default subagents use a low-cost model, but you can request specific models inline:\n\n```\n\u003e /fleet ... Use GPT-5.3-Codex to create the migration.\n  Use Claude Opus 4.5 to analyse test coverage gaps.\n```\n\n**Custom agent assignment** — If you have [custom agents](#custom-agents), reference them with `@`:\n\n```\n\u003e /fleet ... Use @test-writer to create comprehensive unit tests for the endpoint.\n```\n\nCopilot will also auto-select custom agents when it determines they're a good fit.\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eTypical workflow — plan → fleet → autopilot\u003c/summary\u003e\n\n1. Press **Shift+Tab** to switch into [plan mode](#autopilot-mode) and create an implementation plan.\n2. Review and refine the plan with Copilot.\n3. Select **\"Accept plan and build on autopilot + /fleet\"** when the plan is complete.\n4. Copilot divides the plan into parallel subtasks and works through them autonomously.\n\n\u003c/details\u003e\n\n\u003e ⚠️ Each subagent interacts with the LLM independently, so using `/fleet` may consume **more premium requests** than a single-agent session. Use `/model` to check the current model and its multiplier.\n\n#### More examples of prompts that trigger parallel work\n\n| What you ask | What the agent delegates |\n|---|---|\n| *\"Do a security audit of the repo\"* | Parallel scans: XSS in frontend, auth in backend, `npm audit` |\n| *\"Refactor the database layer and make sure nothing breaks\"* | One agent refactors, another runs the test suite continuously |\n| *\"Explain how auth works and also how billing works\"* | Two agents read different parts of the codebase at once |\n| `/fleet` *\"Generate tests for all RecipeShare services\"* | Subagent per service, each writing tests independently |\n\n#### Invoke from a prompt file (VS Code)\n\nYou can wire parallel delegation into a reusable [prompt file](#prompt-files) by adding `agent` to the `tools` frontmatter. The agent then spawns child agents whenever the prompt instructions suggest isolated or parallel work.\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003ereview-feature.prompt.md\u003c/code\u003e\u003c/summary\u003e\n\n```markdown\n---\nname: review-feature\ntools: ['agent', 'read', 'search']\n---\nRun a subagent to research the new feature implementation details and return only\ninformation relevant for user documentation. Then summarise the findings.\n```\n\n\u003c/details\u003e\n\n| | |\n|---|---|\n| **Where** | Copilot CLI (`/fleet`) and VS Code (auto-delegation) |\n| **Context** | Each child agent gets its own isolated context window |\n| **No file needed** | Runtime capability (but can be triggered from a prompt file in VS Code) |\n| **Docs** | [Fleet command](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/fleet) · [Subagents (VS Code)](https://code.visualstudio.com/docs/copilot/agents/subagents) · [Comparing CLI Features](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/comparing-cli-features) |\n\n---\n\n### Autopilot Mode\n\nThe CLI's **autopilot mode** lets Copilot work autonomously through a multi-step task without waiting for your input after each step — you give the initial instruction and Copilot keeps going until the task is complete.\n\n\u003e **When you need it:** You have a well-defined task (implement a plan, write a test suite, refactor a module) and want Copilot to work to completion without back-and-forth interaction.\n\n#### How to enter autopilot mode\n\n- **During an interactive session** — press **Shift+Tab** to cycle through modes until you reach autopilot\n- **From the command line** — start with the `--autopilot` flag:\n\n```bash\ncopilot --autopilot --yolo --max-autopilot-continues 10 -p \"Add input validation to all RecipeShare API endpoints\"\n```\n\nCopilot continues autonomously until one of these happens:\n1. The task is complete\n2. A blocking problem occurs\n3. You press **Ctrl+C**\n4. The `--max-autopilot-continues` limit is reached (if set)\n\n#### Permissions\n\nOn entering autopilot, the CLI prompts you to choose permissions:\n\n1. **Enable all permissions** (recommended) — equivalent to `--allow-all` / `--yolo`\n2. **Continue with limited permissions** — auto-denies tool requests that need approval\n3. **Cancel**\n\nYou can grant full permissions mid-session with the `/allow-all` (or `/yolo`) command.\n\n#### Autopilot + Fleet\n\nAutopilot and `/fleet` are independent features that combine well. A common workflow:\n\n1. Use **plan mode** (Shift+Tab) to create a detailed plan\n2. Select **\"Accept plan and build on autopilot + /fleet\"**\n3. Copilot fans out subtasks to parallel subagents *and* continues autonomously\n\n| | |\n|---|---|\n| **Where** | Copilot CLI only (GA as of v1.0, March 2026) |\n| **Toggle** | Shift+Tab during a session, or `--autopilot` flag |\n| **Cost** | Each autonomous continuation step uses premium requests |\n| **Docs** | [Autopilot mode](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/autopilot) |\n\n---\n\n### Copilot Coding Agent\n\nGitHub's autonomous agent that picks up issues, creates branches, and opens PRs — **without a chat session**. It runs on GitHub's cloud infrastructure, not your machine.\n\n\u003e **When you need it:** You have issues in your backlog that you want worked on asynchronously. You assign them; the coding agent works while you sleep and delivers a PR for review.\n\n#### How it works\n\n1. **Assign an issue to `@copilot`** — via label, Mission Control dashboard, or directly in the issue\n2. **GitHub's cloud infrastructure** spins up an agent with your repo context\n3. **The agent works autonomously** — reads the issue, creates a `copilot/*` branch, makes changes, opens a draft PR\n4. **You review the PR** — treat it like any team member's work\n\n#### Prerequisites\n\n- **Copilot coding agent** enabled on the repo (Settings → Copilot → Coding agent)\n- **`copilot-setup-steps.yml`** in `.github/` — defines the agent's environment\n- **GitHub Actions** enabled\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample — \u003ccode\u003ecopilot-setup-steps.yml\u003c/code\u003e\u003c/summary\u003e\n\n```yaml\nsteps:\n  - uses: actions/checkout@v4\n  - uses: actions/setup-node@v4\n    with:\n      node-version: '22'\n  - run: npm ci\n  - run: npm run build\n```\n\n\u003c/details\u003e\n\n\u003e 💡 The coding agent reads your custom instructions and skills — so everything you've configured for interactive Copilot also applies to the autonomous agent.\n\n| | |\n|---|---|\n| **Where** | GitHub cloud infrastructure |\n| **Interaction** | Async only — no chat, results delivered as PRs |\n| **Context** | Each agent runs in full isolation — separate branch, separate PR |\n| **Cost** | Copilot premium requests |\n\n---\n\n### Mission Control\n\n[Mission Control](https://github.blog/changelog/2025-10-28-a-mission-control-to-assign-steer-and-track-copilot-coding-agent-tasks/) is a centralized dashboard for managing multiple coding agents at scale.\n\n\u003e **When you need it:** You're running multiple coding agents across repos and want a single view to assign, monitor, and steer them.\n\n- See all active coding agents across repos in one view\n- Read real-time session logs and code change rationales\n- Steer, pause, or restart agents mid-execution\n- Assign new tasks to agents directly\n\nAvailable on **GitHub.com**, **VS Code Insiders**, and **GitHub Mobile**.\n\n#### Local vs Cloud agents\n\n| | Running Parallel Tasks | Coding Agent + Mission Control |\n|---|---|---|\n| **Runs on** | Your machine | GitHub cloud infrastructure |\n| **Interaction** | Synchronous — in a chat session | Asynchronous — works while you sleep |\n| **Result** | Agent output in terminal/editor | A pull request |\n| **Memory** | Session-only | `copilot-instructions.md` + issue context |\n| **Scale** | Limited by your machine | Multiple agents across repos |\n\n---\n\n### Agentic Workflows\n\nMarkdown-described automations that run inside **GitHub Actions** — agents triage issues, update docs, fix CI failures, and generate PRs, all triggered by repo events.\n\n\u003e **When you need it:** You want your repo to run on autopilot — issues get triaged, docs stay in sync with code, and CI failures get diagnosed automatically.\n\n**How it works:**\n\n1. **Describe the automation** in a markdown workflow file — what you want done on which trigger\n2. **GitHub Actions invokes the coding agent** with your instructions and repo context\n3. **The agent executes** — reading code, making changes, running tests\n4. **Results appear as PRs** — fully reviewable, with diffs, commit history, and CI results\n\n#### Examples of what you can automate\n\n- Auto-triage and label new issues based on content\n- Diagnose and suggest fixes for CI/CD failures\n- Keep documentation in sync when code changes\n- Generate routine dependency update PRs\n- Improve test coverage on schedule\n\n#### Key details\n\n| | |\n|---|---|\n| **Where** | GitHub Actions + Coding Agent |\n| **Trigger** | Repo events (push, issue created, schedule, etc.) |\n| **Result** | Pull requests — always human-reviewed before merge |\n| **Blog post** | [Automate repository tasks with GitHub Agentic Workflows](https://github.blog/ai-and-ml/automate-repository-tasks-with-github-agentic-workflows/) |\n\n---\n\n### Agents Tab\n\nA repo-level tab on GitHub.com for launching, tracking, and reviewing agent tasks — the repo-scoped companion to Mission Control.\n\n\u003e **When you need it:** You want to see all agent activity for a specific repo in one place, launch new tasks, and review agent-generated PRs without leaving the repo.\n\n- Launch agent tasks directly from the repo\n- Track progress of active agents\n- Review agent-generated pull requests\n- See history of completed agent work\n\n| | |\n|---|---|\n| **Where** | GitHub.com — new tab on each repo |\n| **Scope** | Single repo (vs Mission Control which is cross-repo) |\n\n---\n\n## Platform Features\n\n\u003e Built-in intelligence and extensibility that power the entire Copilot ecosystem.\n\n### Agentic Memory\n\nPer-repo persistent memory — Copilot automatically captures patterns, conventions, and relationships as you work, then validates them against current code before reuse. Memories expire after **28 days** and are shared across all Copilot surfaces.\n\n\u003e **When you need it:** You're tired of re-explaining your architecture, conventions, or recent decisions every new session. You want Copilot to remember what it learned last time.\n\n**How it works:**\n\n- Copilot **automatically captures** small pieces of knowledge (\"memories\") as you work — no manual tagging required\n- Memories are **per-repository** and shared across Copilot features (chat, completions, agents, code review, CLI)\n- Before using a memory, Copilot **validates it against the current code** — if the source changed, the memory is dropped\n- Memories **auto-expire after 28 days** to prevent stale advice\n\n#### Enable it\n\n- **Individual (Pro/Pro+):** GitHub Settings → Copilot → enable **Copilot Memory**\n- **Org/Enterprise:** Settings → Copilot → Policies → **Copilot Memory** → enable for members\n- If you belong to multiple orgs, the most restrictive policy applies\n\n#### View \u0026 manage memories\n\n- **Repo Settings → Copilot → Memory** — review and delete stored memories per repo\n- Repo owners can curate memories; automated validation handles the rest\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eExample: what a memory captures\u003c/summary\u003e\n\nAfter a coding agent session where it learns that `src/api/` and `src/types/` must stay in sync:\n\n\u003e *\"When modifying API route handlers in `src/api/`, the corresponding TypeScript interfaces in `src/types/` must be updated to match.\"*\n\nNext session, Copilot automatically applies this knowledge — no re-prompting needed.\n\n\u003c/details\u003e\n\n| | |\n|---|---|\n| **Scope** | Per-repository, persistent across sessions (28-day expiry) |\n| **Validation** | Just-in-time — checked against source code at use time |\n| **Difference from instructions** | Instructions are manually authored; memory is auto-learned |\n| **Docs** | [Enabling and curating Copilot Memory](https://docs.github.com/en/copilot/how-tos/use-copilot-agents/copilot-memory) |\n| **Blog** | [Building an agentic memory system](https://github.blog/ai-and-ml/github-copilot/building-an-agentic-memory-system-for-github-copilot/) |\n\n---\n\n### Copilot Spaces\n\nProject-specific knowledge containers — persistent, shareable collections of context accessible in Codespaces, dev containers, and across Copilot features.\n\n\u003e **When you need it:** You want to package project knowledge (architecture docs, API specs, onboarding guides) into a container that Copilot always has access to, especially for onboarding new team members.\n\n| | |\n|---|---|\n| **Scope** | Per-project, accessible across Copilot surfaces |\n| **Use for** | Onboarding, architecture context, shared team knowledge |\n| **Supports** | Role-based access, public sharing, templates |\n\n---\n\n### Copilot SDK\n\nProgrammatic access to Copilot's capabilities in **Node.js**, **Python**, **Go**, and **.NET**. Build your own agents as code — with execution loops, tool orchestration, multi-model routing, and context management.\n\n\u003e **When you need it:** You want agents as *code* rather than *prompts*. Or you need to integrate Copilot into custom automation, CI/CD pipelines, or your own developer tools.\n\n```bash\n# Install (Node.js example)\nnpm install @github/copilot-sdk\n```\n\n| | |\n|---|---|\n| **Status** | Technical preview |\n| **Languages** | Node.js, Python, Go, .NET |\n| **Repository** | [github.com/github/copilot-sdk](https://github.com/github/copilot-sdk) |\n| **Difference from agents** | Agents are markdown prompts; SDK is programmatic code |\n\n---\n\n## Community Projects\n\n\u003e Open-source projects built on top of Copilot's features.\n\n### Squad\n\nA persistent team of AI agents with identity, memory, and parallel coordination — now with its own CLI, interactive shell, SDK, and community documentation site.\n\n**Repository:** [github.com/bradygaster/squad](https://github.com/bradygaster/squad) · **Docs:** [bradygaster.github.io/squad](https://bradygaster.github.io/squad/)\n\n```bash\nnpm install --save-dev @bradygaster/squad-cli\nnpx squad init\n```\n\n\u003e **Note:** GitHub-native distribution (`npx github:bradygaster/squad`) has been removed. All distribution is now via npm. See the [Migration Guide](https://github.com/bradygaster/squad/blob/main/docs/get-started/migration.md) if upgrading from an older version.\n\n\u003cdetails markdown\u003e\n\u003csummary\u003eWhat gets created\u003c/summary\u003e\n\n```\n.squad/                            # Team state (committed to git)\n├── team.md                        # Roster — who's on the team\n├── decisions.md                   # Shared brain — team decisions\n├── routing.md                     # Who handles what\n├── ceremonies.md                  # Sprint ceremonies config\n├── casting/\n│   ├── policy.json                # Universe theme \u0026 agent count\n│   ├── registry.json              # Persistent name registry\n│   └── history.json               # Who was cast when\n├── agents/\n│   ├── Keaton/\n│   │   ├── charter.md             # Lead — identity, expertise\n│   │   └── history.md             # What Keaton knows about YOUR project\n│   ├── McManus/charter.md         # Frontend specialist\n│   ├── Verbal/charter.md          # Backend specialist\n│   ├── Fenster/charter.md         # Tester\n│   └── Kobayashi/charter.md       # Scribe\n├── skills/                        # Learned domain knowledge\n├── sessions/                      # Crash recovery checkpoints\n├── log/                           # Session history\n└── orchestration-log/             # What was spawned and why\n\n.github/agents/squad.agent.md     # The coordinator\n```\n\n\u003c/details\u003e\n\n**How it works:** You say *\"Team, build the login page\"* → the coordinator spawns agents in parallel using the `task` tool → each agent works in its own context → decisions are merged via a drop-box pattern → knowledge persists in `history.md` across sessions.\n\n#### Interactive shell \u0026 CLI\n\nRun `squad` with no arguments to enter the interactive shell:\n\n```\nsquad \u003e @Keaton, analyze the architecture of this project\nsquad \u003e Build the login page\nsquad \u003e /status\n```\n\nKey CLI commands: `squad init`, `squad status`, `squad triage` (auto-triage issues), `squad copilot` (add/remove @copilot), `squad doctor`, `squad nap` (context hygiene), `squad export`/`import`, `squad aspire` (observability dashboard).\n\n#### What's new (v0.8.x)\n\n- **SubSquads** — break large teams into focused sub-groups (renamed from workstreams)\n- **Crash recovery** — sessions persist to disk; agents resume from checkpoint after failures\n- **Plugin marketplace** — `squad plugin marketplace add|browse|list`\n- **Azure DevOps adapter** — Squad for enterprise via `CommunicationAdapter`\n- **Upstream sources** — `squad upstream add|sync` to pull from shared squad configs\n- **Context hygiene** — `squad nap --deep` to compress and prune accumulated context\n- **Ralph** — event-driven monitoring agent that watches all agent activity\n\n| | Vanilla Custom Agent | Squad |\n|---|---|---|\n| **Memory** | Fresh every session | `history.md` per agent, `decisions.md` shared |\n| **Identity** | Generic | Persistent themed names (MCU, Usual Suspects, etc.) |\n| **Coordination** | Manual | Auto-routes based on `routing.md` |\n| **Parallelism** | DIY | Built-in fan-out with orchestration logging |\n| **Crash recovery** | None | Persistent sessions with checkpoint resumption |\n| **Git-native** | Just a markdown file | Full `.squad/` folder — clone = get the team |\n\n---\n\n### awesome-copilot\n\nA curated collection of Copilot resources, customizations, and advanced patterns — actively maintained, with community contributions reviewed daily.\n\n**Repository:** [github.com/github/awesome-copilot](https://github.com/github/awesome-copilot)\n\nIncludes custom instructions examples, agent patterns, orchestration strategies, MCP configs, community skills, and guides for maximizing agentic workflows. A great starting point for seeing how others use these features in practice.\n\n---\n\n## Best Practices\n\n\u003e Tips for writing effective instructions, prompts, skills, and agents.\n\n### General Principles\n\n- **Be specific, not vague.** \"React 19 with Vite and Tailwind CSS\" beats \"a React app.\" Name exact frameworks, versions, and tools.\n- **Show, don't just tell.** One concrete code example is worth more than a paragraph of description.\n- **Set boundaries explicitly.** List files, directories, and configs the AI should *never* touch.\n- **Front-load commands.** Put relevant CLI commands near the top — agents act on these first.\n- **Only include what can't be inferred.** Don't duplicate what linters or CI already enforce.\n- **Version control everything.** Check all customization files into your repo.\n\n### Writing Instructions\n\n- Keep under ~1,000 lines — beyond that, context dilution reduces effectiveness\n- Use headings and bullet points, not prose\n- Make rules actionable: ✅ `Never use 'any' type in Typescript ` — ❌ `Try to use good types`\n- Use `applyTo:` globs to scope. Don't put backend rules in a file that applies to frontend\n\n### Writing Prompts\n\n- One prompt = one task. Don't combine unrelated steps\n- Include expected inputs and outputs\n- Use numbered steps for multi-step workflows\n- Specify output format if it matters\n\n### Writing Skills\n\n- Each skill gets its own folder with a lowercased, hyphenated name\n- Include YAML frontmatter with `name` and `description`\n- Write step-by-step with gotchas and edge cases\n- Bundle related scripts in the same folder\n- Clearly specify \"When to use\" so Copilot loads it at the right time\n\n### Writing Agents\n\nBased on [lessons from 2,500+ repositories](https://github.blog/ai-and-ml/github-copilot/how-to-write-a-great-agents-md-lessons-from-over-2500-repositories/):\n\n- **Define a specialist, not a generalist.** \"You are a security reviewer for a Node.js API\" — not \"You are a helpful assistant\"\n- **Cover:** commands/tooling, testing, project structure, code style, git workflow, boundaries\n- **Include positive AND negative instructions.** What to do *and* what to never do\n- **Show output examples.** Demonstrate what a good response looks like\n- **Keep personas focused.** Don't mix frontend and backend concerns — make separate agents\n\n### Authoring Resources\n\n| Resource | Source |\n|---|---|\n| [How to write a great agents.md — Lessons from 2,500+ repos](https://github.blog/ai-and-ml/github-copilot/how-to-write-a-great-agents-md-lessons-from-over-2500-repositories/) | GitHub Blog |\n| [Using custom instructions to unlock the power of Copilot](https://docs.github.com/en/copilot/tutorials/use-custom-instructions) | GitHub Docs |\n| [Creating agent skills](https://docs.github.com/en/copilot/how-tos/use-copilot-agents/coding-agent/create-skills) | GitHub Docs |\n| [Prompt files vs Instructions vs Agents](https://gist.github.com/burkeholland/435ab18c549ddbefde1846165e8b2e08) | Burke Holland |\n| [Mastering GitHub Copilot Customisation](https://onlyutkarsh.com/posts/2026/github-copilot-customization/) | Utkarsh Shigihalli |\n| [All About Copilot Custom Instructions](https://www.nathannellans.com/post/all-about-github-copilot-custom-instructions) | Nathan Nellans |\n\n---\n\n## Quick Reference\n\n| Feature | What | Location | When to use |\n|---------|------|----------|-------------|\n| **Instructions** | Always-on project context | `.github/copilot-instructions.md` | When Copilot doesn't know your stack |\n| **Path-specific** | Per-folder/file overrides | `.github/instructions/*.instructions.md` | When parts need different rules |\n| **Prompts** | Reusable task recipes | `.github/prompts/*.prompt.md` | When you repeat multi-step tasks |\n| **Skills** | Packaged domain knowledge | `.github/skills/{name}/SKILL.md` | When procedures are complex |\n| **MCP Servers** | External tool connections | `.vscode/mcp.json` | When agents need databases/APIs |\n| **Hooks** | Lifecycle automation | `.github/hooks/` or `~/.copilot/hooks/` | When you want auto-formatting/linting/startup prompts |\n| **Agents** | Named specialist personas | `.github/agents/{name}.agent.md` | When you need dedicated workflow owners |\n| **Plugins** | Bundled agent toolkits | `plugin.json` manifest | When sharing agent setups across repos |\n| **Running Parallel Tasks / `/fleet`** | Auto-delegates or explicitly fans out work to subagents | *(runtime capability)* / `/fleet` slash command | When your request has multiple independent jobs |\n| **Autopilot Mode** | Hands-off autonomous task completion | Copilot CLI (`--autopilot` / Shift+Tab) | When you want Copilot to work to completion without interaction |\n| **Coding Agent** | Autonomous cloud agent | GitHub infrastructure | When you want async issue automation |\n| **Agentic Workflows** | AI + GitHub Actions automation | GitHub Actions | When you want automated repo maintenance |\n| **Mission Control** | Multi-agent dashboard | GitHub.com / VS Code | When managing agents at scale |\n| **Agents Tab** | Repo-level agent UI | GitHub.com | When tracking agent work per repo |\n| **Agentic Memory** | Persistent per-repo memory | Built-in | When you want Copilot to remember across sessions |\n| **Copilot Spaces** | Project knowledge containers | GitHub.com | When packaging shared context for teams |\n| **Copilot SDK** | Programmatic agent building | npm / pip / go | When you need agents as code, not prompts |\n\n### File Structure\n\n```\n.github/\n├── copilot-instructions.md              # Global instructions\n├── copilot-setup-steps.yml              # Coding agent environment\n├── instructions/\n│   ├── frontend.instructions.md         # Path-specific\n│   ├── backend.instructions.md\n│   └── tests.instructions.md\n├── prompts/\n│   ├── new-endpoint.prompt.md           # Reusable recipes\n│   └── new-component.prompt.md\n├── skills/\n│   ├── drizzle-migrations/              # Domain knowledge\n│   │   └── SKILL.md\n│   └── recipe-image-upload/\n│       └── SKILL.md\n├── agents/\n│   └── security-reviewer.agent.md       # Custom agents\n├── hooks/\n│   └── hooks.json                       # Repo-level lifecycle hooks\n└── workflows/\n    └── ...\n\n.vscode/\n└── mcp.json                             # MCP server config\n\n~/.copilot/\n├── hooks/                               # Personal hooks (all repos)\n│   └── hooks.json\n└── settings.json                        # CLI user config (enabledPlugins, etc.)\n```\n\n---\n\n## Further Reading\n\n**Official Documentation**\n- [Custom Instructions](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/add-custom-instructions) · [Custom Agents](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/create-custom-agents-for-cli) · [Plugins](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/plugins-creating) · [Comparing CLI Features](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/comparing-cli-features) · [MCP Servers](https://docs.github.com/en/copilot/how-tos/provide-context/use-mcp/extend-copilot-chat-with-mcp) · [Coding Agent](https://docs.github.com/en/copilot/concepts/agents/coding-agent) · [Fleet command](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/fleet) · [Autopilot mode](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/autopilot)\n\n**VS Code**\n- [MCP Servers](https://github.com/microsoft/vscode-docs/blob/main/docs/copilot/customization/mcp-servers.md) · [Subagents](https://code.visualstudio.com/docs/copilot/agents/subagents) · [Agent Skills](https://code.visualstudio.com/docs/copilot/customization/agent-skills) · [Prompt Files](https://code.visualstudio.com/docs/copilot/customization/prompt-files)\n\n**Blog Posts \u0026 Guides**\n- [Maximize Agentic Capabilities](https://github.blog/ai-and-ml/github-copilot/how-to-maximize-github-copilots-agentic-capabilities/) · [Mission Control](https://github.blog/changelog/2025-10-28-a-mission-control-to-assign-steer-and-track-copilot-coding-agent-tasks/) · [Agents vs Skills vs Instructions](https://github.com/orgs/community/discussions/183962) · [Agentic Workflows](https://github.blog/ai-and-ml/automate-repository-tasks-with-github-agentic-workflows/)\n\n**CLI Release Notes**\n- [GitHub Copilot CLI releases](https://github.com/github/copilot-cli/releases) — full changelog for every CLI version (v1.0+ is GA)\n\n**Platform**\n- [Copilot SDK](https://github.com/github/copilot-sdk) · [Copilot Spaces](https://docs.github.com/en/copilot/how-tos/provide-context/use-copilot-spaces)\n\n**Community**\n- [awesome-copilot](https://github.com/github/awesome-copilot) · [Squad](https://github.com/bradygaster/squad) · [Squad Docs](https://bradygaster.github.io/squad/)\n\n---\n\n\u003cdiv align=\"center\"\u003e\n\n**Found this useful?** Give it a ⭐ and share it with your team.\n\n**Something missing?** [Open an issue](https://github.com/nishanil/copilot-guide/issues) or submit a PR.\n\n\u003c/div\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnishanil%2Fcopilot-guide","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnishanil%2Fcopilot-guide","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnishanil%2Fcopilot-guide/lists"}