Desenvolvimento orientado a especificações
Definição
Spec-driven development constrói sistemas de IA (agents, pipelines, tools) a partir de especificações explícitas: requirements, output formats, allowed actions, and constraints. Specs are retrieved and used at runtime (por ex. in RDD) so behavior stays aligned with intent.
É especially useful for agents and RDD: instead of encoding all rules in weights or prompts, you maintain specs (por ex. in docs or a knowledge base) and retrieve them at runtime. Fits regulated domains and teams that want behavior to be auditable and updatable sem retreinar.
Como funciona
Você escreve especificações (linguagem natural, esquemas ou regras estruturadas) e as indexa para recuperação (por ex. em um armazenamento vetorial store or structured repo). At runtime, the task (and optionally the current state) is used to retrieve relevant spec fragments. The model or agent decides (por ex. next step, allowed actions) and generates (output, tool call) with the spec in context. Validate checks the output against the spec (por ex. schema, rules); if validation fails, you can retry or surface an error. This keeps generation and decisãos aligned with the spec without baking everything into prompt engineering or fine-tuning.
Casos de uso
Spec-driven development fits when behavior must stay aligned with retrievable requirements (RDD, compliance, or safety).
- Building agents that retrieve and follow specs (por ex. RDD pattern)
- Enforcing output format and constraints (JSON, allowed actions)
- Regulated or safety-critical flows where behavior must match requirements
Documentação externa
- LangChain – Structured output — Enforcing output format from LLMs
- OpenAI – Structured outputs