bmad-spec

To generate machine-readable software requirements, you can distill inputs from PRDs and RFCs into a standardized spec kernel. This process outputs a five-field structure covering capabilities, constraints, and success signals for clear downstream communication.

A spec kernel structures software development intent into a standardized machine-readable format. It organizes inputs into five distinct fields: Why, Capabilities, Constraints, Non-goals, and Success signal, ensuring clear communication for downstream tasks.

Yes, you can convert a Game Design Document into a standardized spec kernel. The tool distills requirements from various project documentation formats, including GDDs, transforming vague inputs into structured machine-readable outputs.

Yes, you need Python 3 to process specification kernels and companion files. Python execution is required for applying templates and resolving customizations during the transformation of intent inputs into the final machine-readable format.

To automate spec generation for programmatic integration, you can use the headless operation mode. This allows you to process inputs and generate spec kernels without interactive guidance, enabling seamless integration into automated pipelines.

A limitation of using generated spec kernels is that they focus strictly on five core fields, requiring optional companion files like glossaries and diagrams for additional detail. Complex project management documentation may need supplementary processing.