Ahmad Awais

Canada

Multiple Award-winning Open Source GitHub Star ❯ Founder & CEO ⌘ Command Code most used coding agent for open models 250T tokens ❯ Ex VP of DevTools/DX Rapid Google Developers Advisory Board (gDAB) founding member ❯ Node.js Community Committee Outreach Lead ❯ Course author at VSCode.pro & NodeCLI.com ❯ Ex VP of Engineering DevRel WGA ❯ WordPress Core Team Dev ❯ EE-CS Engineer

Community Contributions

LatestBlog postsOpen sourceVideo/PodcastSpeaking ArticlesEvent organizationHackathonOther

GitHub Copilot App: Contributed Shades of Purple code theme in the GitHub App

One of my most important contributions to GitHub is the Shades of Purple Code theme, purpose-built for the GitHub Copilot app, which is now live, both in dark and light variants. Shades of purple theme historically has always been a dark theme, but for GitHub Copilot app, I was asked to build a lighter version, which was a heavy-duty task, I won't lie. But I'm glad to see it live. and give over ten million developers who use this theme the ability to read the code and enjoy the github app with the coolest shades of purple.

Other / 08-01-2026

how did /design fix the ai design slop problem

Most AI-generated interfaces look the same. /design is the part of Command Code that knows why, and fixes it in your files.

Video/Podcast / 06-25-2026

how did we make deepseek outperform opus 4.7?

how did we make deepseek outperform opus 4.7? i've been thinking about why "open model bad at tool calling" is almost always a harness problem, not a model problem. especially as i build https://commandcode.ai context: spent the two days looking at billions of tokens in ‪@CommandCodeAI‬ (tb open source ai cli) using deepseek. I ended up writing a tool-input repair layer. the trigger was watching deepseek-flash fail on the simplest /review run, every shellCommand and readFile call bouncing back with a raw zod issues blob, the model unable to recover because the error wasn't in a form it could read. by the end deepseek v4 pro was beating opus 4.7 6/10 times on our internal evals.

Video/Podcast / 06-18-2026

i may be open sourcing Command Code, this is why!

This is one of the biggest product decisions I’m thinking through right now as the founder of Command Code. I’ve spent 25+ years building open source software, contributing across ecosystems like React, Node, WordPress, VS Code, Chrome, Bootstrap, NASA Mars Ingenuity tooling, MCP, and more. Open source is a big part of who I am as a developer. But Command Code is also a company. We recently raised $5M from incredible investors. I have employees to take care of, investors to report to, and a mission to build the best AI engineering developer experience in the world. So the question is not that simple. Open sourcing Command Code could make it more hackable, extensible, transparent, and community-driven. It could let developers build their own workflows, plugins, extensions, and taste-driven coding environments on top of what we’re building.

Speaking (conference/usergroups) / 05-03-2026

How I beat Claude & GPT with Kimi in just $1 with my coding taste using Command Code

In this video I build a full TypeScript CLI from scratch using Kimi K2.5 (a Sonnet-level model at 4x the speed and 25x less cost), demo how Command Code's "taste" system auto-learns your coding preferences per project, and stress test all three open source models on a real build. By the end I had a working CLI, a taste file that knew my entire stack (pnpm, tsup, vitest, clack, pico colors), and I'd spent exactly $0.74 of my $1 plan.

Video/Podcast / 04-29-2026

I can't believe Cloudflare actually shipped this

Command Code with your coding taste https://CommandCode.ai use open source models and continuous learning to beat Claude and GPT.

Video/Podcast / 04-14-2026

GitHub Actions: Building CommandCode.ai Coding Agent

Introducing Command a coding agent powered by Langbase, Claude, OpenAI, GitHub Models and orchestrated with GitHub Actions. Ahmad Awais demonstrates building AI agents with meta neuro-symbolic reasoning that automate PR reviews, code analysis, and refactoring directly in your GitHub repos.

Speaking (conference/usergroups) / 10-15-2025

GitHub Actions x Copilot x Langbase: Building & Running Coding Agents

Discover how to orchestrate intelligent coding agents that work while you sleep. Ahmad Awais demonstrates building autonomous agents using GitHub Actions for workflow automation, GitHub Copilot for context-aware code generation, and Langbase for persistent memory and knowledge retrieval. Learn to create agents that review PRs, refactor code, generate documentation, and handle repetitive dev tasks—all running serverless in your CI/CD pipeline. This talk delivers practical patterns for composing these primitives into production-ready agents that scale across your entire codebase.

Event organization / 10-02-2025

Related Stars

Jessica Temporal

jtemporal

Nick Taylor

nickytonline

Rishab Kumar

rishabkumar7

Adrian Hajdin

adrianhajdin