The concept of spec-driven development has gained popularity as a solution to the issue of AI agents generating code that is subtly incorrect. This approach involves writing a structured spec and letting the AI agent execute against it, with the goal of reducing errors. Several tools and frameworks, such as Spec Kit, OpenSpec, BMAD, and Kiro, have been built on this idea. However, the assumption that a spec can remain the source of truth is flawed, as it requires human effort to keep it updated. The real challenge in this space is creating specs that can update themselves without human intervention. Spec-driven development has become the default answer to the problem of AI agents generating incorrect code. Many companies and organizations are investing in this approach, with GitHub Spec Kit having over 90,000 stars and Tessl raising $125 million in funding. Despite the popularity of spec-driven development, the issue of keeping specs up to date remains an unsolved problem. The current workflow for spec-driven development involves writing a spec, planning, and implementing, but this process can be time-consuming and prone to errors. The development of specs that can update themselves automatically could potentially revolutionize the field of software development and address the limitations of current spec-driven development approaches.