How-toSelected

Runley agents now write nearly all Runway App Studio PRs

Why we picked it

The non-obvious takeaway is that full agent-authored codebases are feasible when review and testing are automated and humans are reserved for risk-gated decisions. Engineering leads evaluating AI-assisted development workflows can use this as a reference architecture for their own CI pipelines. Teams across ASEAN and global markets building internal developer platforms will find the browser-testing pattern directly reusable.

Runway says that since the start of October, agents have written nearly every pull request merged into Runway App Studio, its AI-powered app-building product. The claim comes from a new engineering post published on runway.com/news/engineering that explains the system behind this workflow.

Image · Runway ML
Body

The system routes agent-generated code through an automated pipeline: changes are reviewed programmatically, tested inside a browser environment, and escalated to a human engineer only for high-risk decisions. This lets the team merge agent-written PRs while keeping human oversight where it matters most, such as changes with security, data, or user-facing consequences.

The post is notable because Runway, a company best known for Gen-3 and its video generation models, is applying agentic coding at scale to its own product development. It offers a concrete reference implementation for teams experimenting with agent-driven software engineering, showing how automated review and browser-based testing can substitute for traditional code review at most risk levels.

Original · Runway ML

Since the start of October, agents have written nearly every pull request we’ve merged into Runway App Studio. Our latest engineering post explains how we built a system that reviews those changes, tests them in a browser and brings in an engineer for high-risk decisions. runway.com/news/engineeri…📊 10 ⚡

Industry Coordinates & Source Verification

This report is categorized under AITOP's "How-to" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Runway, agents, coding assistants, workflows.

Source attribution: curated from Runway ML (@runwayml).