Best of AI-Assisted Development — 2025

  1. 1
    Article
    Avatar of hnHacker News·42w

    Read That F*cking Code!

    AI coding tools like Claude Code enable developers to generate working code without reading it, but this practice leads to three critical issues: architectural decay, loss of domain knowledge, and security vulnerabilities. The author advocates for two responsible approaches: fast prototyping with post-session review for peripheral features, and synchronous pair-coding for core functionality. A comprehensive checklist covers architecture consistency, security scoping, meaningful tests, documentation, error handling, and performance considerations.

  2. 2
    Article
    Avatar of itnextITNEXT·47w

    Hacker Screen Builder

    A web-based tool that creates cinematic hacker-style interfaces through drag-and-drop widgets including satellite maps, 3D globes, progress bars, and facial recognition animations. Built using AI assistance from ChatGPT for ideation, Google AI Studio for frontend generation, and GitHub Copilot for backend development. Features customizable widgets, screen testing, and shareable links for saved configurations.

  3. 3
    Video
    Avatar of mattpocockMatt Pocock·1y

    Build An MCP Server In 5 Prompts // Vibe Coding That Doesn't Suck

    Learn how to build an MCP server in Typescript using AI-assisted development in just five prompts. This guide covers setting up the server, integrating it with GitHub using OctoKit, converting it to SSE with Express, and structuring your codebase effectively. The emphasis is on planning, documentation, and leveraging AI to streamline development tasks.

  4. 4
    Video
    Avatar of TechWithTimTech With Tim·1y

    This Free AI Coding Assistant Might Destroy Cursor...

    Augment Code is a new AI coding assistant designed to integrate with VS Code and various JetBrains IDEs. It excels in managing and making productive changes in larger codebases rather than building projects from scratch. Key features include fast context-aware code generation, instant code syncing, and comprehensive indexing. It offers a free version, which allows unlimited usage but trains on user data, and paid versions that secure user code. Augment Code is noted for its speed and efficiency compared to other AI assistants like GitHub Copilot and Wind Surf.

  5. 5
    Article
    Avatar of hnHacker News·44w

    How Anthropic teams use Claude Code

    Anthropic's internal teams across 10 departments share how they use Claude Code to transform their development workflows. Teams report significant productivity gains through automated code generation, debugging assistance, and cross-functional collaboration. Key use cases include infrastructure debugging, test generation, codebase navigation, and enabling non-technical staff to build complex applications. Teams emphasize the importance of detailed documentation, iterative workflows, and treating Claude Code as a collaborative partner rather than a one-shot solution.

  6. 6
    Video
    Avatar of galaxies_devSimon Grimm·26w

    I Built a Mobile Farming Game in 6 Days (with AI + Expo)

    A developer shares their experience building a mobile farming game in six days using React Native, Expo, and AI tools. The workflow involved using ChatGPT for planning, Cursor for coding, and various AI models for different tasks. Key learnings include leveraging Expo Application Services (EAS) for automated builds and deployments, the challenges of game design versus development, and the limitations of AI when debugging complex state management issues. The developer emphasizes that while AI dramatically accelerated development, understanding the underlying technology remains essential for solving critical bugs.

  7. 7
    Article
    Avatar of portkeyportkey·42w

    Claude Code vs Cursor: What to choose?

    Claude Code and Cursor represent two distinct approaches to AI-assisted coding. Claude Code is a terminal-based assistant that excels at complex, one-shot tasks with deep codebase understanding and automation capabilities, but comes with higher costs and a steeper learning curve. Cursor is an AI-enhanced VS Code fork that provides familiar IDE experience with visual feedback and multiple interaction options, making it more beginner-friendly but sometimes less reliable with large files. The choice depends on your workflow preferences, team size, budget, and specific use cases.

  8. 8
    Article
    Avatar of aspnetASP.NET Blog·30w

    Introducing Custom Agents for .NET Developers: C# Expert & WinForms Expert

    Microsoft released two experimental custom agents for GitHub Copilot targeting .NET developers: C# Expert and WinForms Expert. C# Expert provides guidance on modern C# best practices, code integrity, and testing workflows. WinForms Expert specializes in UI design patterns, event-driven programming, and protects designer.cs files from corruption. Both agents can be downloaded from the awesome-copilot repository and integrated into projects via the .github/agents folder, with support across Copilot Coding Agent, VS Code Insiders, and upcoming Visual Studio 2022 v17.14.21.

  9. 9
    Article
    Avatar of logrocketLogRocket·39w

    I tried out Kiro: Here’s what I learned

    Kiro is AWS's new AI-powered IDE that introduces a spec-driven development workflow, requiring developers to define requirements and designs before code generation. Unlike traditional AI coding tools that rely on vague prompts, Kiro structures the development process through three phases: requirements, design, and implementation planning. The author tested Kiro by building a retro Tetris game and adding multiplayer functionality, finding that the structured approach produces more reliable results than typical 'vibe coding' sessions. Kiro also features agent hooks for automating repetitive tasks and positions itself more as an engineering assistant than a simple coding copilot.

  10. 10
    Article
    Avatar of on_leuwcqSamuel Adekunle·28w

    Vibe Coding: Build A Complete App From Scratch In Minutes Using GitHub Copilot.

    A senior engineer shares their journey from skepticism to adoption of GitHub Copilot after testing it on tutorial projects. They demonstrate building a Flutter app called "Mood Mixer" that captures selfies, detects emotions, and generates Spotify playlists using only AI prompts and no manual coding. The piece explores how AI tools can serve as effective assistants for scaffolding, state management, and function improvements rather than complete replacements for developer skills.

  11. 11
    Article
    Avatar of ghblogGitHub Blog·34w

    Spec-driven development: Using Markdown as a programming language when building with AI

    A developer shares an experimental approach to building applications by writing specifications entirely in Markdown and using GitHub Copilot to compile them into actual code. The workflow involves maintaining app logic in markdown files, using AI to generate Go code, and keeping documentation synchronized with implementation. This spec-driven development method aims to solve context loss issues with AI coding agents while enabling faster iteration and cleaner specifications.

  12. 12
    Article
    Avatar of ergq3auoeReinier·37w

    CodeBuff: ClaudeCode KILLER! New AI Coding Agent is Quite Powerful, FREE, & Opensource!

    CodeBuff is a new open-source, multi-agent AI coding assistant that offers both CLI and SDK integration for development workflows. It positions itself as a powerful alternative to existing AI coding tools like ClaudeCode, providing free access to developers.

  13. 13
    Article
    Avatar of github_updatesGitHub Changelog·52w

    Introducing Copilot Spaces: A new way to work with code and context

    GitHub introduces Copilot Spaces, a new feature that centralizes project context including code, documentation, and custom instructions to make Copilot more knowledgeable about specific projects. Teams can create shared spaces to distribute expertise across organizations, with files automatically staying up-to-date from repositories. The feature is currently in early preview for all Copilot users, with billing changes taking effect June 4th, 2025.

  14. 14
    Article
    Avatar of vibecodingVibe Coding·34w

    Vibe coding just got an upgrade with Claude Sonnet 4.5

    Anthropic released Claude Sonnet 4.5, their latest and most capable coding model that reportedly outperforms the previous Opus 4.1 version.

  15. 15
    Article
    Avatar of devopsDevOps.com·35w

    Hugging Face Opens GitHub Copilot Chat to Open-Source Models

    Hugging Face released a VS Code extension that integrates open-source large language models directly into GitHub Copilot Chat. Developers can now access hundreds of inference providers and switch between specialized models like DeepSeek V3.1, GLM 4.5, and Kimi K2 without leaving their editor. The integration offers flexibility to choose models optimized for specific programming languages, frameworks, or domains, breaking free from proprietary model limitations. The extension requires VS Code 1.104.0 or newer and offers a free tier for experimentation with pay-as-you-go pricing for additional capacity.

  16. 16
    Article
    Avatar of zedZed·49w

    Agentic Engineering in Action — Zed's Blog

    Mitchell Hashimoto demonstrates his approach to AI-assisted development while building Ghostty terminal emulator. He treats AI tools like junior engineers, providing clear architectural guidance and well-scoped problems rather than open-ended tasks. His workflow involves maintaining architectural control, working in parallel with AI agents for tasks like refactoring and cleanup, and using multiple AI models competitively. While AI excels at refactoring and simple bug fixes, it struggles with architectural decisions, high-performance data structures, and languages like Zig.

  17. 17
    Article
    Avatar of collectionsCollections·46w

    OpenCode: An Advanced Open-Source AI Coding Agent for Terminal

    OpenCode is an open-source AI coding agent designed for terminal use, offering access to over 75 AI models from providers like OpenAI, Anthropic, and Google Gemini. It features collaborative workflows with sharable session links, parallel agent processing, customizable themes, and superior performance compared to alternatives like Gemini CLI and ClaudeCode. The tool supports multi-agent and multi-session capabilities, making it a versatile coding assistant for developers.

  18. 18
    Article
    Avatar of devtoDEV·23w

    My OSS Stalled for 3 Months Because of Misguided Vibe Coding—This Is the Full Reboot Story

    A developer shares their experience recovering from a 3-month project stall on diffx, a structured-data diff tool. The root causes included monorepo complexity, over-sharing CI/CD infrastructure across multiple projects, premature multilingual documentation, and quality issues from AI-assisted "vibe coding" without proper verification. The recovery process involved quarantining existing code, manually verifying functionality, writing specs from scratch, rebuilding tests based on specs rather than implementation, and splitting the monorepo into separate language-specific repositories. Key lessons include the importance of spec-driven development with AI, two-stage release workflows, and having the courage to delete broken artifacts rather than clinging to sunk costs.

  19. 19
    Article
    Avatar of googledevsGoogle Developers·27w

    Introducing the Jules extension for Gemini CLI

    Google launched the Jules extension for Gemini CLI, an autonomous coding assistant that handles tasks asynchronously in a background virtual machine. While Gemini CLI serves as an interactive terminal collaborator for active tasks, Jules works independently to clone code, install dependencies, and modify files. Developers can delegate tasks using the /jules command and check progress without interrupting their workflow. The extension requires Gemini CLI v0.4.0 or newer and can be installed with auto-update capabilities.

  20. 20
    Article
    Avatar of kiroKiro·32w

    The wait(list) is over, get started with Kiro today

    Kiro, an AI-powered IDE from AWS, has removed its waitlist and is now publicly available. New users receive 500 free bonus credits (50% of the Pro plan) valid for 30 days. Version 0.4.0 introduces spec-driven development features including optional MVP tasks, per-prompt credit consumption visibility, dev server integration, and the ability to reference specs as context. The IDE uses a unified credit system that meters usage in 0.01 increments, with different models consuming credits at varying rates.

  21. 21
    Article
    Avatar of kiroKiro·29w

    This is Kiroween

    Kiro announces Kiroween, a Halloween-themed hackathon with $100,000 in prizes across 12 categories. Participants build applications using Kiro's agentic IDE features including specs, agent hooks, steering, and MCP. The competition runs from October 31 to December 5, 2025, with categories like Resurrection, Frankenstein, Skeleton Crew, and Costume Contest, plus a special $10,000 startup prize. All participants receive Kiro Pro+ tier access during the submission period.

  22. 22
    Article
    Avatar of phProduct Hunt·41w

    stagewise: The frontend coding agent for existing codebases

    stagewise is an open-source frontend coding agent that runs in your browser and works with existing local codebases. It allows developers to visually edit UI elements by clicking on them and using natural language prompts like 'increase the height here' to make changes directly to source code. The tool bridges the gap between visual prototyping tools like v0 and traditional local development workflows, supporting any frontend framework and integrating with localhost development servers.

  23. 23
    Article
    Avatar of windsurfWindsurf·22w

    Windsurf Wave 13: Merry Shipmas

    Windsurf Wave 13 introduces parallel agents for concurrent task execution, Git worktrees support, multi-pane Cascade interface, a dedicated terminal, and SWE-1.5 Free tier. These updates enhance the AI-powered coding assistant's capabilities for managing complex development workflows.

  24. 24
    Article
    Avatar of vsVisual Studio Blog·31w

    Introducing Planning in Visual Studio (Public Preview)

    Visual Studio 2022 17.14 introduces Planning in Chat, a public preview feature that enables GitHub Copilot to handle complex, multi-step coding tasks through structured plans. Instead of simple prompts, Copilot now creates markdown-based plans that break down large tasks, research codebases, and execute steps iteratively while adapting to new context. The feature uses transparent tool calls and hierarchical planning techniques, showing 15% higher success rates in SWE-bench testing. Plans are stored as temporary files with progress tracking, allowing developers to monitor and control each step of the AI-assisted workflow.

  25. 25
    Article
    Avatar of github_updatesGitHub Changelog·35w

    Claude Opus 4.1 is now generally available in GitHub Copilot

    Claude Opus 4.1 is now available in GitHub Copilot for Pro+ and Enterprise users across multiple IDEs including VS Code, Visual Studio, JetBrains, Xcode, and Eclipse. Enterprise administrators need to enable the new policy in Copilot settings, while Pro+ users can activate it through the model picker with a one-time confirmation prompt.

See all AI-Assisted Development archives