Spec-driven development frameworks we track: how each one works, which tools it supports, how mature it is, and which real projects use it. Compare them side by side, or, if one is missing, submit it.
Skill-based agent discipline with hard gates: a mandatory brainstorm → spec → plan pipeline with human approval before any code, then per-task subagent execution under strict test-driven development.
Matt Pocock's composable family of agent skills built around "grilling" — a relentless one-question-at-a-time interview that aligns human and agent before code is written — chaining into spec, tickets, and TDD implementation.
Open-source toolkit from GitHub that scaffolds a phase-gated, spec-first workflow into any of 40+ AI coding agents.
Change-folder-based SDD for AI coding assistants — fluid, iterative, and explicitly built for brownfield codebases, not just greenfield.
Agile AI development framework that simulates a full team of specialized agent personas across a four-phase lifecycle, scaling its ceremony to project complexity.
Context-engineering-first SDD loop — fresh-context subagents, persistent state artifacts, and phase-gated milestones — continued by the community OpenGSD org after the original repo was archived.
Standards-first system that extracts your codebase's real conventions and injects them into agent context to shape better specs.
AWS's harness-neutral implementation of a gated, adaptive development life cycle spanning intent, requirements, implementation, deployment, and operational feedback.
Repository-native SDD plugin that carries durable product context into a track-based spec, plan, implementation, review, and recovery loop.
Kiro-style spec harness installed as Agent Skills into eight coding agents from one npx command, now centered on long-running autonomous implementation with per-task independent review.
Spec Kit descendant that extends spec, plan, and task generation into a governed, repository-native execution loop with work packages, isolated worktrees, review state, acceptance, and merge decisions.
Reverse-engineering-first SDD framework that turns legacy code into confidence-scored, traceable specifications before an agent evolves, migrates, debugs, or refactors the system.
Requirements-centric SDD that adapts the Unified Process for AI: use cases owned by requirements engineers, with code, tests, and docs regenerated around them.
Maximalist open-source SDD: EARS-format requirements, a nine-article constitution, CLI-enforced requirements-to-test traceability, and delta specs, installed into seven coding agents from one npx command.
A complete SDLC for specifications, grounded in Agile and Test-Driven Development.
Spec-driven development for Claude Code in which specs, rather than chat history, are the control surface, and generated projects carry tracked Git hooks that enforce closed-spec discipline on pushes to the trunk.
Augment Code's commercial "Unified Agents Platform": a reviewed spec as the human checkpoint for fleets of agents — Experts with persistent memory and durable Sessions, orchestrated org-wide on Augment's Context Engine.
AWS's agentic IDE where specs are a native feature: requirements, design, and task artifacts with approval gates, living in .kiro/ inside the repo.
A JetBrains IDE plugin that makes any AI coding agent spec-driven — write a spec and plan, then execute it step by step with a review gate after every step, without leaving the IDE.
Commercial "spec-as-source" platform from Snyk founder Guy Podjarny: a .spec.md framework, a versioned registry of evaluated agent context, and MCP integration into any coding agent.