<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr" -->

---
title: VS Code Learn: Agent-First Development | daily.dev
description: A multi-part video series teaches agent-first development using VS Code and GitHub Copilot. It covers the five ingredients for effective agent use (harness,...
canonical: https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: VS Code Learn: Agent-First Development | daily.dev
og:description: A multi-part video series teaches agent-first development using VS Code and GitHub Copilot. It covers the five ingredients for effective agent use (harness,...
og:url: https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr
og:image: https://api.daily.dev/og/posts/uY4z0hRSr.png
og:image:alt: VS Code Learn: Agent-First Development
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# VS Code Learn: Agent-First Development

**[Visual Studio Code](https://daily.dev/sources/vscode)** · 48 min read · 4 upvotes · 0 comments

## Summary

A multi-part video series teaches agent-first development using VS Code and GitHub Copilot. It covers the five ingredients for effective agent use (harness, model, context, tools, prompts), the different approval/permission levels (default, bypass, autopilot), reviewing and steering agent work, forking sessions and restoring checkpoints, managing multiple sessions locally, in the cloud, or via Copilot CLI, and using agent debug logs and the chat debug view to inspect tool calls, token usage, and context window consumption. The final video ties everything together by building a URL shortener from scratch using plan mode and autopilot.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://www.youtube.com/watch?v=K040jc0PQWI>

## Questions this post answers

### What is the difference between default approvals, bypass approvals, and autopilot mode in GitHub Copilot agent mode in VS Code?

Default approvals uses configured settings and asks before running commands; bypass approvals auto-approves all tool calls and retries but still asks the agent to pause for clarification questions; autopilot (in preview) auto-approves all tool calls and answers clarification questions itself to push the task through to completion without stopping. Autopilot is set from the same approval dropdown used for default and bypass.

_Developers deciding how much autonomy to grant coding agents can compare workflow write-ups like this on daily.dev._

### What are the five things needed for a successful agent session in VS Code, according to GitHub Copilot's agent-first workflow?

The five components are harness, model, context, tools, and prompts. The harness (VS Code plus Copilot chat) wires together model access, tool calls, and file access; the model does reasoning at a chosen thinking effort (low, medium, high); context includes files and instructions provided or read; tools let the agent act; and prompts direct the task, together forming a repeatable formula for reliable agent output.

_Anyone refining their agent prompting workflow can find similar practical breakdowns on daily.dev._

### How can I check how much of my GitHub Copilot agent's context window is being used and what's consuming it?

The context window indicator in VS Code's Copilot chat shows a percentage (turning red as it fills) and breaks usage into system instructions, tool definitions, user context (messages and responses), and tool results such as terminal output or file reads. You can manually trigger summarization with the compact conversation option or the slash-compact command to reduce token usage, such as a 200,000-token window example shown at 10% usage.

_Developers troubleshooting bloated agent sessions can track similar context-management tips on daily.dev._

---

Tags: [#llm](https://daily.dev/tags/llm), [#github](https://daily.dev/tags/github), [#ai-agents](https://daily.dev/tags/ai-agents), [#mcp](https://daily.dev/tags/mcp), [#vscode](https://daily.dev/tags/vscode)

[View this post on daily.dev](https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"VS Code Learn: Agent-First Development","url":"https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr"},"datePublished":"2026-09-17T17:41:21.688Z","dateModified":"2026-09-17T18:07:12.384Z","description":"A multi-part video series teaches agent-first development using VS Code and GitHub Copilot. It covers the five ingredients for effective agent use (harness,...","image":"https://i.ytimg.com/vi/K040jc0PQWI/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/K040jc0PQWI/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Visual Studio Code","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Visual Studio Code","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/d5adab6ae9fb44d19859748fc30f5e18","url":"https://daily.dev/sources/vscode"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,github,ai-agents,mcp,vscode","timeRequired":"PT48M","video":{"@type":"VideoObject","name":"VS Code Learn: Agent-First Development","description":"A multi-part video series teaches agent-first development using VS Code and GitHub Copilot. It covers the five ingredients for effective agent use (harness,...","thumbnailUrl":"https://i.ytimg.com/vi/K040jc0PQWI/sddefault.jpg","uploadDate":"2026-09-17T17:41:21.688Z","duration":"PT48M","url":"https://api.daily.dev/r/uY4z0hRSr","embedUrl":"https://www.youtube.com/embed/K040jc0PQWI"}}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Visual Studio Code","item":"https://daily.dev/sources/vscode"},{"@type":"ListItem","position":3,"name":"VS Code Learn: Agent-First Development"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/vs-code-learn-agent-first-development-uy4z0hrsr#faq","mainEntity":[{"@type":"Question","name":"What is the difference between default approvals, bypass approvals, and autopilot mode in GitHub Copilot agent mode in VS Code?","acceptedAnswer":{"@type":"Answer","text":"Default approvals uses configured settings and asks before running commands; bypass approvals auto-approves all tool calls and retries but still asks the agent to pause for clarification questions; autopilot (in preview) auto-approves all tool calls and answers clarification questions itself to push the task through to completion without stopping. Autopilot is set from the same approval dropdown used for default and bypass. Developers deciding how much autonomy to grant coding agents can compare workflow write-ups like this on daily.dev."}},{"@type":"Question","name":"What are the five things needed for a successful agent session in VS Code, according to GitHub Copilot's agent-first workflow?","acceptedAnswer":{"@type":"Answer","text":"The five components are harness, model, context, tools, and prompts. The harness (VS Code plus Copilot chat) wires together model access, tool calls, and file access; the model does reasoning at a chosen thinking effort (low, medium, high); context includes files and instructions provided or read; tools let the agent act; and prompts direct the task, together forming a repeatable formula for reliable agent output. Anyone refining their agent prompting workflow can find similar practical breakdowns on daily.dev."}},{"@type":"Question","name":"How can I check how much of my GitHub Copilot agent's context window is being used and what's consuming it?","acceptedAnswer":{"@type":"Answer","text":"The context window indicator in VS Code's Copilot chat shows a percentage (turning red as it fills) and breaks usage into system instructions, tool definitions, user context (messages and responses), and tool results such as terminal output or file reads. You can manually trigger summarization with the compact conversation option or the slash-compact command to reduce token usage, such as a 200,000-token window example shown at 10% usage. Developers troubleshooting bloated agent sessions can track similar context-management tips on daily.dev."}}]}
```

