मुख्य सामग्री पर जाएं

skills · GitHub Copilot as a coding agent

GitHub Copilot & AI Coding Assistants

Use GitHub Copilot's agent on a real repo, and decide what it may do.

कोर्स के अंत तक, आप ये कर पाएँगे:

  • A first agent session verified against the files
  • A test-first fix with its diff reviewed
  • Approval settings you chose on purpose
  • Instructions files and a plan before a multi-file change
  • A custom agent, a code review and an MCP decision
  • A planted-text test, a sandbox decision and a practiced undo
  • A CLI script, a reviewed cloud-agent pull request and a teammate note

$9/माह से7.5 घंटे5 अध्याय20 पाठ

कोर्स अभी अंग्रेज़ी में हैं।

“Copilot can open the pull request now. Who reads it?”

इस कोर्स के बारे में

यह किसके लिए है

  • You code daily and want Copilot's agent on a repo that matters, not a demo
  • You use Copilot completions or chat and want to know exactly what the agent reads, runs and edits
  • You need a Copilot setup a teammate can adopt: instructions, approvals and checks in the repo

आपको क्या चाहिए

  • You code and use git and GitHub
  • VS Code and a GitHub account (Copilot Free is enough to start)

आप कौन से कौशल अभ्यास करेंगे

  • GitHub Copilot
  • Copilot agent mode
  • Custom instructions
  • Custom agents
  • Copilot code review
  • MCP
  • Copilot CLI
  • Copilot cloud agent

कितना समय लगता है

कुल मिलाकर लगभग 7.5 घंटे

  1. अध्याय 1 · Copilot is an agent now1.5 घंटे
  2. अध्याय 2 · Context is the job1.5 घंटे
  3. अध्याय 3 · Make it yours1.5 घंटे
  4. अध्याय 4 · What it must never do1.5 घंटे
  5. अध्याय 5 · Beyond the editor1.5 घंटे

इसमें क्या शामिल है

  • 20 पाठ
  • 81 पठन
  • 20 हल किए गए उदाहरण
  • 40 अभ्यास
  • 40 त्वरित जाँचें
  • 20 आम भ्रम
  • 20 क्विज़
  • अध्याय के अंत की 5 जाँचें

पाठ्यक्रम

5 अध्याय, 20 पाठ

पूरा कोर्स: हर अध्याय और पाठ। विस्तार के लिए अध्याय पर टैप करें। शुरू करने पर पाठ खुलेंगे।

    • Instructions It Reads Every Time
    • What It Can See: # Mentions and Search
    • Plan Before a Big Change
    • Give It a Way to Check Its Work
    • A Custom Agent for a Task You Repeat
    • Choosing the Model and What It Costs
    • A Second Look: Copilot Code Review
    • Connect One Tool with MCP
    • Text It Reads Can Give It Orders
    • Secrets, the Sandbox and Approvals That Are Not Boundaries
    • Plausible Is Not Correct
    • Undo: Checkpoints, Git and Branches
    • Run It From the Terminal: The Copilot CLI
    • Hand an Issue to the Cloud Agent
    • Several Agents at Once
    • The Note Your Teammate Needs

करियर

यह कोर्स आपको कहाँ तक ले जा सकता है

वे भूमिकाएँ जिनसे यह कोर्स जुड़ता है, और वे कौशल जिन्हें कोर्स पूरा करने के बाद आप अपने रिज़्यूमे और LinkedIn पर लिख सकते हैं।

वे कौशल जो आप बनाएँगे

कोर्स पूरा करने के बाद इन्हें अपने रिज़्यूमे या LinkedIn प्रोफ़ाइल में चिपकाएँ।

  • Running GitHub Copilot's agent in VS Code and checking its answersUse GitHub Copilot agent mode in VS Code
  • Choosing between Copilot inline suggestions, Ask, Plan and AgentUse AI code completion and chat assistants appropriately for the task
  • Test-first bug fixing with an AI coding agentWrite a failing test before a fix and review the resulting diff
  • Configuring Copilot agent approvals in VS CodeConfigure tool approval and auto-approve rules for an AI coding agent
  • Writing repository custom instructions for GitHub CopilotConfigure custom instructions for an AI coding assistant
  • Giving an AI coding agent the right contextProvide relevant code context to AI assistants
  • Planning multi-file changes with Copilot's Plan agentReview and edit AI-generated implementation plans
  • Verifying AI-generated changes with tests, linters and command outputVerify AI-generated code with automated tests and linters
  • Writing custom agents for GitHub CopilotCreate reusable AI agent workflows for a codebase
  • Managing GitHub Copilot model choice and usage costsMonitor and manage AI tool usage and spend
  • Reviewing changes with Copilot code reviewUse AI-assisted code review on pull requests
  • Connecting MCP servers to GitHub Copilot safelyConfigure Model Context Protocol servers for AI agents
  • Testing an AI coding agent for prompt injectionAssess AI agent exposure to prompt injection
  • Keeping secrets out of reach of AI coding agents and configuring the agent sandboxManage secrets safely around AI developer tools
  • Reviewing AI-generated code against the specificationReview AI-generated pull requests for correctness
  • Recovering from AI agent changes with checkpoints and gitUse version control to roll back AI-generated changes
  • Running GitHub Copilot CLI in scripts with scoped tool permissionsAutomate developer tasks with an AI coding agent from the command line
  • Delegating issues to a cloud coding agent and reviewing its pull requestsWrite issues with acceptance criteria for AI coding agents
  • Running parallel AI agent sessions with git worktreesIsolate concurrent AI agent work in separate git worktrees
  • Documenting team practice for an AI coding assistantWrite team guidelines for using AI coding assistants in a repository

प्रश्न

प्रतिबद्ध होने से पहले

You write code, know git and GitHub pull requests, and want Copilot's agent to work in a real repo of yours while you still decide what it reads, runs and merges.

Most of it. As of November 2026, signing in to VS Code without a plan enrolls an eligible account in Copilot Free, whose agents are limited, and every plan includes the Copilot CLI. The cloud agent and code review on pull requests need a paid plan or Copilot Student; those lessons give Free users a local path or a written decision instead. GitHub's plans page has the current details.

This course does not promise it. The often-quoted 55% comes from a 2022 GitHub experiment with 95 developers writing one HTTP server in JavaScript, using the inline completions of that time, not today's agent. Lesson 4.3 reads that study closely, and you track your own results.

No. The invented sample repo is a short Python project because it runs anywhere, but the required step in each lesson is on your own repo, in whatever language it uses.

Use the invented sample repo; every lesson has a path through it. A repo of your own gets more out of the course, as long as you may send its code to an AI service.

आज ही शुरू करें