GitHub Was Built for Human Developers. What Happens When Agents Take Over?

A lot can go wrong in seven hours and 47 minutes.

On August 17, GitHub suffered a worldwide outage that took down repositories, pull requests, GitHub Actions, APIs, authentication and Copilot. If you were planning to ship software that Monday, there was a decent chance you did something else instead.

Apparently, a critical infrastructure component failed to scale when traffic hit a new peak. GitHub knew it was bad. "We let you down," the company wrote afterward.

Things were about to get worse, though.

In the middle of GitHub's very bad day, Cursor released Origin, an early-beta code-hosting platform featuring repositories, pull requests, code browsing, integrations with deployment and continuous integration tools, and GitHub syncing.

In other words, Cursor released a starter pack of the things developers have spent the better part of two decades relying on GitHub for. The timing was almost comically bad.

One day later, on August 18, Warp launched Factories, infrastructure for what it calls AI software factories: systems of agents that can work through the stages of software development with greater autonomy and less human intervention.

And then, on August 20, Salesforce introduced Slack Code, allowing teams to work with coding agents, including GitHub Copilot, directly inside Slack.

All of that happened in four days.

For a moment, it looked like GitHub was facing a particularly uncomfortable question: If AI agents are going to write, review and deploy more of our software, what exactly do we still need GitHub for?

Then GitHub started answering.

At GitHub Universe in late October, the company unveiled a wave of agent-focused capabilities, including tools for coordinating multiple agents, automating development tasks and choosing models for different jobs.

And on November 6, GitHub announced that its own agent integration in Slack and Microsoft Teams was generally available.

The race is no longer just about who has the best coding assistant. It's about who controls the environment where agents do their work.

And GitHub, it turns out, has no intention of surrendering that position.

The first generation of AI coding tools fit neatly into existing development processes.

GitHub was where the code lived. Your integrated development environment was where you worked on it. Maybe an AI assistant helped you write a little faster.

Then the assistants got ambitious.

Rather than suggesting the next few lines of code, tools like Cursor and Claude Code started behaving like agents. They can search codebases, edit multiple files, run commands, test changes, fix problems and keep going.

Suddenly, infrastructure designed around human developers began looking negotiable. Cursor made that negotiation explicit with Origin.

For now, it's a friendly negotiation. Developers don't have to break up with GitHub. Origin can mirror GitHub repositories in real time. Developers can review a GitHub pull request from Cursor, comment on it, merge it and keep GitHub as the source of truth.

The consequential part is that Cursor also lets developers host repositories directly on Origin, without GitHub in the path.

Cursor is building an alternative place for code, agents and collaboration to live together, not merely offering a more convenient interface for working with GitHub. The distinction matters because the company that controls the workflow has more opportunities to shape everything that happens around it.

And Cursor isn't the only company rethinking where software development should happen.

Salesforce's Slack Code takes a different approach.

Instead of moving developers into another coding environment, it brings coding agents into the place where teams are already having conversations.

Tag an agent in Slack, and a team can work with it in a dedicated code channel. The agent's plan, changes and previews become visible to the people who need to review or redirect the work.

Say a product manager notices a bug and brings it up in Slack. An engineer tags an agent, which investigates the issue, proposes a fix and generates a preview. The team reviews the result and requests changes without switching between half a dozen applications.

The code may still live on GitHub, but the humans increasingly don't need to spend much time there. And here's where the story gets interesting: GitHub is embracing that same idea.

On November 6, GitHub announced that @GitHub in Slack and Microsoft Teams was generally available. Teams can ask GitHub's agent to investigate bugs, create issues, implement changes and open pull requests directly from a conversation.

So Slack isn't replacing GitHub but it is another interface through which GitHub operates. That's a very different competitive picture from the one we saw in August. The battle is about which company gets to own the interaction between people, agents and code.

Warp Factories pushes the idea further.