aidd-methodology

AI-driven development requires context-first documentation that anchors architecture, contracts, and limitations upfront. You create ADRs and design briefs to establish a single source of truth, ensuring LLM agents receive unambiguous guidance before implementation begins.

Planning complex multi-service projects with LLM agents requires upfront architectural guidance and design briefs to coordinate cross-team development. Establishing review checkpoints before implementation ensures the agent operates within defined constraints and architectural boundaries.

ADRs and design briefs provide the context-first requirements and architectural guidance that LLM agents need for implementation. They create a single source of truth that eliminates ambiguity, allowing developers to act as architects who review agent output against documented constraints.

Context-first documentation supports cross-team iteration planning by maintaining updated architecture and contract documents for agent-based workflows. It ensures all teams reference a single source of truth during documentation updates and subsequent implementation cycles.

AI-driven development requires clear delineation of roles where developers act as architects providing oversight, while LLM agents handle implementation. The architect must review context-first documentation and establish checkpoints before empowering the agent to write code.

Upfront architectural documentation for LLM projects is less suited for simple, single-service scripts or rapid prototypes where architectural constraints and cross-team coordination are unnecessary. It targets complex projects requiring strict context-first planning and contract definition.