When agents sped up, two teams hit the same wall: validation couldn't keep up
Agents made it faster to write code. Two engineering teams published what happened on the validation side when they did, and the two write-ups land on the same structural claim from opposite ends. Linear found its CI gate became the bottleneck. Salesforce found its human review stopped engaging. Neither team concluded that generation was the problem. Both concluded that the part of the system…
Two engineering teams published their experiences when implementing AI-generated code, revealing that validation couldn't keep up with the increased speed. Linear found their CI gate became the bottleneck, while Salesforce encountered issues with human review engagement. Neither team identified generation as the problem; instead, they discovered the verification system couldn't scale with the faster code generation.
Linear reworked their validation machinery by using faster third-party runners, optimizing specific compiler processes, and replacing custom lint rules with AST-based linting. Salesforce, on the other hand, rebuilt their review system around reconstructing intent, treating context as a first-class dependency, and progressively disclosing risk to reviewers.
Both teams addressed the same underlying issue: the verification system couldn't keep pace with the faster code generation, highlighting the need for teams to identify which end of the validation funnel is breaking for them.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
