Skip to content

How GitHub uses AI agents to stress test new features

Read the Portuguese translation →

Author: GitHub · Watch on YouTube

Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.

Original language: English

Original video: How GitHub uses AI agents to stress test new features · Channel: GitHub · 00:30 · Watch on YouTube

Topics: AI agents, Cybersecurity and software security, Software engineering with AI

The speakers describe using a swarm of AI agents to stress test a feature by trying to break it and circumvent its rules. They say the agents ran around the clock and proved crafty and resourceful. The conversation frames agent-driven testing as a new type of testing, building on test-driven development. The agents draw on human creativity and are relentless in trying to find ways to break the feature.

  • The team used a swarm of agents to test a feature.
  • The agents ran around the clock.
  • They tried to break the feature and circumvent its rules.
  • The speakers describe the agents as crafty and resourceful.
  • The conversation calls agent-driven testing a new type of testing.
  • The speakers contrast this approach with test-driven development.
  • The agents are described as relentless.
Start Chapter Summary
00:00 Agents put the feature to the test The speakers describe using a swarm of agents to break the feature and circumvent its rules.
00:30 A new era of testing The speakers characterize agent-driven testing as a new era, with agents drawing on human creativity and relentlessly trying to break things.

we just had a swarm of agents that were just running around the clock — 00:00

trying to break the feature, trying to circumvent the rules — 00:00

I think it’s it’s a new type of testing. — 00:00

they are they are relentless. — 00:30

GitHub, agent-driven testing, test-driven development, swarm of agents

00:00 It’s fascinating to have to think about all of these different use cases where it’s like how can someone break our pristine system? >> And you know what’s even better nowadays is you can have agents do this as well. That’s something that we did actually quite extensively is we just had a swarm of agents that were just running around the clock trying to break the feature, trying to circumvent the rules, do all sorts of things. >> That’s a good way to test. >> very crafty, they’re very resourceful, and agent-driven testing. I think it’s it’s a new type of testing. We’ve always

00:30 had test-driven development, but now I think there’s this new era of have the agents go at it and try and break it. And they have learned from all our creativity, and they are they are relentless.