How Developers can Use AI in the Terminal

If you think "vibe coding" is just fancy copy-paste from ChatGPT, you're not doing it right. When I demo CLI-based vibe coding, jaws hit the floor. The difference isn't in the chat window, it's in the terminal where your code assistant becomes your full-stack orchestrator. I'm talking about Claude Code or AWS Q for Developer integrated with your entire ecosystem: AWS CLI, GitHub, Linear, Docker, local services! Anthropic launched Claude Code in research preview February 2025, going fully live with Claude 4 in May 2025. OpenAI followed with their Codex CLI in April 2025. Google joined the party with Gemini CLI in July 2025. AWS had been quietly building this capability through their Q for Developer platform, evolving from CodeWhisperer. The CLI became the new battleground for AI-assisted development. Your CLI, whether on your local machine or in the cloud, coupled with CLI tools to external services like GitHub and AWS, plus MCP services to Linear, gives your code assistant access to everything in your terminal. You can deploy an EC2 instance without knowing the syntax. But here's the workflow that blows minds: with the right prompts you can watch it pull story details from Linear, write the code in VSCode, run the tests in Docker, generate a descriptive commit message, push to your remote repo, create a pull request, and then update the Linear issue with the PR link and status change to "In Review." That's a complete development cycle executed by describing intent in plain English. Watch this approach: spin up multiple terminal windows with different git branches for the same feature. Have your assistant try different approaches across those branches simultaneously - one exploring a React solution, another testing a Vue approach, maybe a third experimenting with server-side rendering. You've just multiplied your development resources and can compare real working code. Just make sure to use descriptive branch names (feature/react-approach, feature/vue-approach) and clean up the unused branches afterward to avoid repo clutter. That's like having a senior developer who can actually execute across your whole infrastructure stack. They're not just suggesting docker commands or AWS deployment steps, they're running them. Building your app, spinning up containers locally, pushing to cloud services, deploying to production environments - all while you focus on the business logic. I don't need to context-switch between my IDE, terminal, AWS console, and project management tools. The assistant handles the orchestration layer while I stay in flow state. It's not about memorizing complex commands anymore - it's about describing intent and watching it happen. This is where AI-assisted development gets genuinely transformative. We're not just automating code generation, we're automating the entire development workflow.

AI coding assistants are changing the way software gets built. I've recently taken a deep dive into three powerful AI coding tools: Claude Code (Anthropic), OpenAI Codex, and Cursor. Here’s what stood out to me: Claude Code (Anthropic) feels like a highly skilled engineer integrated directly into your terminal. You give it a natural language instruction, like a bug to fix or a feature to build and it autonomously reads through your entire codebase, plans the solution, makes precise edits, runs your tests, and even prepares pull requests. Its strength lies in effortlessly managing complex tasks across large repositories, making it uniquely effective for substantial refactors and large monorepos. OpenAI Codex, now embedded within ChatGPT and also accessible via its CLI tool, operates as a remote coding assistant. You describe a task in plain English, it uploads your project to a secure cloud sandbox, then iteratively generates, tests, and refines code until it meets your requirements. It excels at quickly prototyping ideas or handling multiple parallel tasks in isolation. This approach makes Codex particularly powerful for automated, iterative development workflows, perfect for agile experimentation or rapid feature implementation. Cursor is essentially a fully AI-powered IDE built on VS Code. It integrates deeply with your editor, providing intelligent code completions, inline refactoring, and automated debugging ("Bug Bot"). With real-time awareness of your codebase, Cursor feels like having a dedicated AI pair programmer embedded right into your workflow. Its agent mode can autonomously tackle multi-step coding tasks while you maintain direct oversight, enhancing productivity during everyday coding tasks. Each tool uniquely shapes development: Claude Code excels in autonomous long-form tasks, handling entire workflows end-to-end. Codex is outstanding in rapid, cloud-based iterations and parallel task execution. Cursor seamlessly blends AI support directly into your coding environment for instant productivity boosts. As AI continues to evolve, these tools offer a glimpse into a future where software development becomes less about writing code and more about articulating ideas clearly, managing workflows efficiently, and letting the AI handle the heavy lifting.

I just open-sourced a system that 10x'd my development speed with Claude Code. It's called Claude Code Orchestrator, and it enables parallel AI development — multiple Claude sessions working simultaneously on different parts of your codebase. The Problem When you use AI coding assistants sequentially, you're constantly waiting. Build authentication... wait 10 minutes. Now build the API... wait another 10. Write tests... wait again. Or try to do multiple at the same time and deal with merge conflicts / compaction loss / context rot. The Solution What if you could spawn 5 Claude sessions at once, each working on a different piece of the puzzle, without merge conflicts? This doesn't make everything perfect, but it makes parallel dev a lot easier. That's exactly what this does: /spawn auth "implement user authentication" /spawn api "create REST API endpoints" /spawn tests "write comprehensive test suite" Each worker runs in:  • Its own terminal (iterm2) tab  • Its own git worktree (isolated directory)  • Its own feature branch Zero merge conflicts. True parallelism. The Automation Layer The real magic is the orchestrator loop. Start it and walk away:  • Workers get initialized automatically  • CI status is monitored  • Code reviews run via built-in QA agents  • PRs auto-merge when all checks pass  • Finished workers clean themselves up I've been running 10+ parallel workers on complex features while focusing on architecture decisions. Built-in Quality Gates Every PR passes through specialized agents before merge:  • QA Guardian — code quality and test coverage  • DevOps Engineer — infrastructure review  • Code Simplifier — cleans up large changes (from Boris Cherny, creator of Claude Code) Try It One command to install: curl -fsSL https://lnkd.in/gAmsFhtT | bash Requirements: macOS with iTerm2 This is based on patterns from Boris Cherny, creator of Claude Code at Anthropic. The future of software development isn't AI replacing developers — it's developers orchestrating fleets of AI workers. GitHub: https://lnkd.in/gTp6wjy7  #AI #SoftwareDevelopment #Productivity #OpenSource #Claude #Anthropic

Most AI tools slow you down. Not because the AI is bad, but because it drags you out of flow. You end up jumping between browsers, editors, and random scripts just to do simple stuff like tests, reviews, or docs. And that extra friction adds up. I’ve been trying something different: A terminal‐first AI. 👉 Qodo Gen CLI No new UIs. No context switching. Just a few commands wired into my existing setup. Here’s what makes it click for me: • Covers the SDLC: tests, code reviews, changelogs, release notes. • It supports any AI model you choose • Fits into your existing tools, scripts, and CI/CD. • Gives you automation you can trust, all triggered from your terminal. No new UIs. No constant context-switching; it is part of your setup. What’s one thing you’d want an AI to do in your terminal every day?

𝗧𝗵𝗲 𝘁𝗲𝗿𝗺𝗶𝗻𝗮𝗹 𝗶𝘀 𝗯𝗲𝗰𝗼𝗺𝗶𝗻𝗴 𝘁𝗵𝗲 𝗻𝗲𝘄 𝗽𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴 𝗜𝗗𝗘—𝗮𝗴𝗮𝗶𝗻. It’s the simplest, but most powerful interface. And all frontier model vendors now offer CLI versions of their coding agent tools. On my OpenClaw dev machine, I don’t bother installing VS Code, Cursor, or other IDEs (besides the pre-installed Xcode). I work in the terminal with Codex, Claude Code, and GitHub Copilot. Recently, I used Claude Code to design, plan, and implement an entirely new messaging feature for Teams as a POC idea. But it didn’t handle @mentions correctly through the Graph API—so I asked Codex to debug and fix it, and it worked like a charm. With multiple coding agents orchestrated on the same machine, it feels like the old pair-programming experience—but with much more productivity and capability. The real ceiling is how many agents one person can manage without burnout. Andrej Karpathy recently posted on X reflecting this paradigm shift in programming: “𝘐𝘵 𝘪𝘴 𝘩𝘢𝘳𝘥 𝘵𝘰 𝘤𝘰𝘮𝘮𝘶𝘯𝘪𝘤𝘢𝘵𝘦 𝘩𝘰𝘸 𝘮𝘶𝘤𝘩 𝘱𝘳𝘰𝘨𝘳𝘢𝘮𝘮𝘪𝘯𝘨 𝘩𝘢𝘴 𝘤𝘩𝘢𝘯𝘨𝘦𝘥 𝘥𝘶𝘦 𝘵𝘰 𝘈𝘐 𝘪𝘯 𝘵𝘩𝘦 𝘭𝘢𝘴𝘵 𝟤 𝘮𝘰𝘯𝘵𝘩𝘴: 𝘯𝘰𝘵 𝘨𝘳𝘢𝘥𝘶𝘢𝘭𝘭𝘺 𝘢𝘯𝘥 𝘰𝘷𝘦𝘳 𝘵𝘪𝘮𝘦 𝘪𝘯 𝘵𝘩𝘦 “𝘱𝘳𝘰𝘨𝘳𝘦𝘴𝘴 𝘢𝘴 𝘶𝘴𝘶𝘢𝘭” 𝘸𝘢𝘺, 𝘣𝘶𝘵 𝘴𝘱𝘦𝘤𝘪𝘧𝘪𝘤𝘢𝘭𝘭𝘺 𝘵𝘩𝘪𝘴 𝘭𝘢𝘴𝘵 𝘋𝘦𝘤𝘦𝘮𝘣𝘦𝘳. 𝘛𝘩𝘦𝘳𝘦 𝘢𝘳𝘦 𝘢 𝘯𝘶𝘮𝘣𝘦𝘳 𝘰𝘧 𝘢𝘴𝘵𝘦𝘳𝘪𝘴𝘬𝘴 𝘣𝘶𝘵 𝘪𝘮𝘰 𝘤𝘰𝘥𝘪𝘯𝘨 𝘢𝘨𝘦𝘯𝘵𝘴 𝘣𝘢𝘴𝘪𝘤𝘢𝘭𝘭𝘺 𝘥𝘪𝘥𝘯’𝘵 𝘸𝘰𝘳𝘬 𝘣𝘦𝘧𝘰𝘳𝘦 𝘋𝘦𝘤𝘦𝘮𝘣𝘦𝘳 𝘢𝘯𝘥 𝘣𝘢𝘴𝘪𝘤𝘢𝘭𝘭𝘺 𝘸𝘰𝘳𝘬 𝘴𝘪𝘯𝘤𝘦 - 𝘵𝘩𝘦 𝘮𝘰𝘥𝘦𝘭𝘴 𝘩𝘢��𝘦 𝘴𝘪𝘨𝘯𝘪𝘧𝘪𝘤𝘢𝘯𝘵𝘭𝘺 𝘩𝘪𝘨𝘩𝘦𝘳 𝘲𝘶𝘢𝘭𝘪𝘵𝘺, 𝘭𝘰𝘯𝘨-𝘵𝘦𝘳𝘮 𝘤𝘰𝘩𝘦𝘳𝘦𝘯𝘤𝘦 𝘢𝘯𝘥 𝘵𝘦𝘯𝘢𝘤𝘪𝘵𝘺 𝘢𝘯𝘥 𝘵𝘩𝘦𝘺 𝘤𝘢𝘯 𝘱𝘰𝘸𝘦𝘳 𝘵𝘩𝘳𝘰𝘶𝘨𝘩 𝘭𝘢𝘳𝘨𝘦 𝘢𝘯𝘥 𝘭𝘰𝘯𝘨 𝘵𝘢𝘴𝘬𝘴, 𝘸𝘦𝘭𝘭 𝘱𝘢𝘴𝘵 𝘦𝘯𝘰𝘶𝘨𝘩 𝘵𝘩𝘢𝘵 𝘪𝘵 𝘪𝘴 𝘦𝘹𝘵𝘳𝘦𝘮𝘦𝘭𝘺 𝘥𝘪𝘴𝘳𝘶𝘱𝘵𝘪𝘷𝘦 𝘵𝘰 𝘵𝘩𝘦 𝘥𝘦𝘧𝘢𝘶𝘭𝘵 𝘱𝘳𝘰𝘨𝘳𝘢𝘮𝘮𝘪𝘯𝘨 𝘸𝘰𝘳𝘬𝘧𝘭𝘰𝘸.” 🔗 [X post link in comment section]

🐙 Let Your AI Fetch from GitHub Directly You're deep in a coding session. You need context from that GitHub issue to fix the bug. So you switch to GitHub, find issue 2013, copy the relevant bits, switch back, and paste. There's a better way. I saw a question in Slack the other day: "Do we have access to the GitHub MCP server?" The person asking wanted their AI coding assistant to fetch from GitHub without the manual shuffle. Reasonable! But here's what clicked for me: if you're using Claude Code, Cursor, or any similar tool with terminal access, you already have what you need. 🔧 Your LLM Can Use the GitHub CLI These tools can run any CLI command you can, including gh, the GitHub CLI. You're not waiting on someone to build or deploy an MCP server. You don't need a specialized integration. You just need to tell your coding assistant what you want from GitHub, and it'll figure out how to get it. Here's the thing: you don't even need to know the specific gh commands. Just ask naturally: > Use the GitHub CLI to fetch the details from github dot com/owner/repo/issues/2013 and use that to... > Use the GitHub CLI to pull in the discussion and diff from github dot com/owner/repo/pull/5678 to understand..." Your assistant will work out which gh commands to run. 🚀 The Pattern Extends Beyond GitHub This isn't just about GitHub. Any CLI tool you have installed becomes a potential integration point for your LLM. Your assistant can use jq for JSON manipulation, sed for text processing; you're not limited by which MCP servers exist or which integrations someone has built. If there's a CLI for it, your LLM can use it. ✨ Try It Next Time Next time you're about to copy-paste context from GitHub into your coding assistant, try this instead: > Use the GitHub CLI to fetch issue github dot com/owner/repo/issues/2013 and use that context to... No need to specify how. Just say what you need and watch what happens. You might never copy-paste from GitHub again. #aitools #developertools #softwareengineering

I think I just found the first AI CLI that’s actually worth using. I’ve been testing Qoder CLI for the past few days, and I didn’t expect it to change how I work this much. If you live in the terminal, Git, SSH, CI/CD, quick fixes, refactoring, this thing feels weirdly natural. What surprised me most is that it isn’t trying to replace your IDE. It brings Qoder’s context engine into the terminal so you can switch between both without losing context, credits, or history. Same account. Same memory. Same agents. Just… faster. Here’s what really stood out for me: → Quest Mode + Worktree This one feels unreal. You can spin up multiple “feature branches” as AI Quests that run in parallel, each with its own isolated workspace. I asked it to fix a caching bug, build a payment module, and refactor an old auth flow… all at once. It felt like having 3 senior engineers working quietly in the background while I focused on design decisions instead of implementation. → Terminal-native speed Most AI CLIs feel heavy. Qoder CLI is insanely light. Responses hit in under 200ms on my machine, and memory usage is tiny. It actually feels like a terminal tool, not a glued-on wrapper. → AI Code Review Quick /review before committing? You get structured feedback directly in the terminal. No switching. No browser tab. And it actually catches things that usually slip in during late-night coding. → Custom commands + sub-agents This part is addictive. You can create your own AI-powered commands with one sentence for your project or globally. I made one that generates test files matching my team’s structure. Another that explains any module’s architecture when I forget how something works. → MCP integration You can plug the CLI into external systems like databases, APIs, and browsers, and call them through natural language. This makes the CLI feel less like “AI autocomplete” and more like an actual engineering agent that can execute real actions. → Long-term memory The AGENTS.md project memory file gets smarter as you work. Sessions restore cleanly. And the main and sub agent setup keeps context from getting polluted. But the biggest shift for me? I didn’t realize how much time I waste jumping between environments. With Qoder CLI, I can stay in the terminal longer but still tap into deep project context and multi-step automation. If you rely on the terminal every day, whether you're doing DevOps, backend, automation, or working across multiple IDEs, this might be worth a look. Try it here: https://aisecret.co/Denis And if you end up testing Quest Mode, be prepared. It’s a productivity rabbit hole. #QoderCLI #DevTools #AIengineering #terminalworkflow #AIagents #productivityforengineers #DenisPanjuta

OpenAI Releases Codex CLI: An Open-Source Local Coding Agent that Turns Natural Language into Working Code OpenAI has introduced Codex CLI, an open-source tool designed to operate within terminal environments. Codex CLI enables users to input natural language commands, which are then translated into executable code by OpenAI’s language models. This functionality allows developers to perform tasks such as building features, debugging code, or understanding complex codebases through intuitive, conversational interactions. By integrating natural language processing into the CLI, Codex CLI aims to streamline development workflows and reduce the cognitive load associated with traditional command-line operations. Codex CLI leverages OpenAI’s advanced language models, including the o3 and o4-mini, to interpret user inputs and execute corresponding actions within the local environment. The tool supports multimodal inputs, allowing users to provide screenshots or sketches alongside textual prompts, enhancing its versatility in handling diverse development tasks. Operating locally ensures that code execution and file manipulations occur within the user’s system, maintaining data privacy and reducing latency. Additionally, Codex CLI offers configurable autonomy levels through the --approval-mode flag, enabling users to control the extent of automated actions, ranging from suggestion-only to full auto-approval modes. This flexibility allows developers to tailor the tool’s behavior to their specific needs and comfort levels...... Read full article here: https://lnkd.in/gDjDNq3u GitHub Repo: https://lnkd.in/gh8nwKCg OpenAI

In the last 4 months, I've tried and tested 7 coding AI Agents Here are my top 5, and here's when to use them as well... Coding agents cannot be ignored now. These agents are not only moving markets but are the core foundation of AI-Native companies in 2026. And in the last 4 months, I tested major coding agents, from open source options like Open Code, Cline, to paid options like Cursor. 📌 Let me break down all 5 coding agents so you get to when to use which one: 1\ OpenAI Codex - Cloud-based coding agent that runs tasks in isolated sandboxes via CLI - Best for: Background/async tasks, parallel agents, CI/CD pipelines - Use when: You need to automate large-scale coding tasks without touching the IDE Quick Start Guide: https://lnkd.in/gKUgFnPH 2\ Claude Code - Anthropic's terminal-based agentic coding tool — works directly in your codebase - Best for: Large refactors, multi-file edits, complex debugging - Use when: You live in the terminal and need deep, repo-level reasoning Quick Start Guide: https://lnkd.in/gSAYPN4b 3\ GitHub Copilot - AI pair programmer embedded across VS Code and the entire GitHub ecosystem - Best for: Inline autocomplete, quick snippets, PR reviews - Use when: You want frictionless suggestions without changing your existing workflow Quick Start Guide:https://lnkd.in/gdmERrn5 4\ Cursor - AI-native code editor (fork of VS Code) with deep codebase understanding - Best for: Complex cross-platform testing, faster edits across multiple files - Use when: You want an AI-first editor that understands your full project context Quick Start Guide: https://lnkd.in/gf8jbtdv 5\ Antigravity - Google's autonomous code editor — agent-first IDE (fork of VS Code) powered by Gemini 3 - Best for: End-to-end task execution, native Google model APIs, browser-based testing - Use when: You want to act as an architect and delegate full tasks to autonomous agents Quick Start Guide: https://lnkd.in/gQXQbHHk 📌 Quick decision guide: 1\ Need async, sandboxed task automation → Codex  2\ Terminal-first, large codebase refactoring → Claude Code  3\ Daily autocomplete within VS Code/GitHub → Copilot  4\ AI-native editor with deep project context → Cursor  5\ Orchestrate multiple agents end-to-end → Antigravity If you want to understand AI agent concepts deeper, my free newsletter breaks down everything you need to know: https://lnkd.in/g5-QgaX4 Save 💾 ➞ React 👍 ➞ Share ♻️ & follow for everything related to AI Agents