August 2025 Summaries

Hatchet faced significant challenges in maintaining accurate documentation for their rapidly evolving SDKs across multiple programming languages, as traditional methods resulted in outdated or non-functional examples, causing drift and inefficiencies. To address this, Hatchet developed a novel system where documentation examples are not separate files but are integrated as living, working code within the SDK repositories, ensuring they are linted, type-checked, and tested in the CI pipeline. This approach, supported by a simple comment-based markup system for code extraction, has led to consistently reliable examples, streamlined updates across languages, and improved developer velocity, with no more broken examples reaching users and providing seamless onboarding experiences. Additionally, Hatchet plans to use LLMs to suggest updates for explanatory text to prevent content drift, enhancing the sustainability of their documentation practices.