Software Methodology
How AI fundamentally changes software delivery paradigms — not just augmenting existing phases, but reshaping the methodology itself.
This domain is distinct from:
- Product Lifecycle entries (which augment a specific phase within the current methodology)
- Horizontal entries (which are lifecycle-agnostic PM skills and tools)
- Adjacent Disciplines (which track how collaborator disciplines adopt AI)
The test: is this about how the methodology itself is evolving rather than how to use AI within the current methodology?
Entries
- Compound Engineering Loop — Four-step loop (Plan → Work → Assess → Compound) where each feature makes the next easier; inverts traditional complexity debt
- OAI Harness Engineering — Zero manually-written code at scale; humans steer (environments, intent, feedback loops), agents execute everything; rigid architecture + mechanical enforcement + continuous entropy management
Related Areas
- Spec-Driven Development — A related methodology shift: software as specs + tests with zero implementation code
- Knowledge Capture as Side Effect — The mechanism that makes the “compound” step work: corrections become persistent institutional memory