Introducing Agentic PRs

author image

Tim Nichols

CEO/Founder

2025-08-13T03:03:26.251Z

tldr - Flightcrew now creates PRs in response to dynamic issues in your code and cloud

Flightcrew creates Agentic PRs

When you’re hit by a traffic spike or vendor outage (or you deploy something incorrect), Flightcrew has your back and will create the code, PR and analysis needed to remediate.

What’s new about Agentic PRs?

We first launched Flightcrew it operated like a dependabot + linter for cloud infrastructure issues. Flightcrew would …

  1. Identify issues in your cloud infrastructure, like poor kubernetes resource usage
  2. Crawl a repo to find the file and line(s) of code responsible
  3. Generate the code/config and analysis to resolve the issue

Crucially, these PRs were metered out based on your declared preferences (ex: Help me resolve compliance drift in X repo every Y days). This scheduled pattern made sense for our initial use cases which tended to be high-volume, recurring tasks.

As we graduate into high-difficulty, high-upside tasks like refactoring IAC, we needed a new event-based architecture to meet our users where they are and catch falling knives in live cloud environments.

Agentic PRs generated dynamically in response to emerging issues in your infrastructure.

How does Flightcrew Prioritize Agentic PRs?

Flightcrew has a global understanding of infrastructure risks, but prioritizes Agentic PRs based on what it knows about your cloud infrastructure.

If you’ve defined service tiering or SLOs in a machine-readable format, or have defined preferences for scheduled PRs, then Flightcrew will use this context to prioritize PRs. If not, then Flightcrew will target PRs based on SRE best practices: prioritizing the reliability of central services over last mile optimizations

Flightcrew will learn from your team’s feedback (approval, rollbacks, comments, emojis, etc) or lack thereof to improve its sense of pacing and priority.

Use Agentic PRs to improve reliability, slash costs, and automate compliance.

We’re incredibly excited by this release both for its immediate value and long-term potential.

Once installed, Flightcrew has your back and does the last-mile of work necessary to make the cloud safe and self-serve. Give it a try!

author image

Tim Nichols

CEO/Founder

Tim was a Product Manager on Google Kubernetes Engine and led Machine Learning teams at Spotify before starting Flightcrew. He graduated from Stanford University and lives in New York City. Follow on Bluesky

keep-reading-vector
Subscription decoration

Don’t miss out!

Sign up for our newsletter and stay connected