157 lines
14 KiB
Markdown
157 lines
14 KiB
Markdown
# Awesome Spec-Driven Development
|
|
|
|
A curated list of spec-driven development (SDD) frameworks, agent workflows, and complementary skills for building software with AI coding agents. SDD keeps intent, requirements, design, tasks, and verification as reviewable artifacts rather than transient chat context.
|
|
|
|
## Contents
|
|
|
|
- [SDD frameworks](#sdd-frameworks)
|
|
- [Workflow and methodology systems](#workflow-and-methodology-systems)
|
|
- [Agent skills by purpose](#agent-skills-by-purpose)
|
|
- [Task, context, and knowledge management](#task-context-and-knowledge-management)
|
|
- [Agent runtimes and orchestration](#agent-runtimes-and-orchestration)
|
|
- [Verification, quality, and security](#verification-quality-and-security)
|
|
- [Protocols and repository conventions](#protocols-and-repository-conventions)
|
|
- [Platforms and hosted tools](#platforms-and-hosted-tools)
|
|
- [Standards and learning](#standards-and-learning)
|
|
|
|
## SDD Frameworks
|
|
|
|
- [OpenSpec](https://github.com/Fission-AI/OpenSpec) - Lightweight, Markdown-first SDD with proposals, specs, designs, tasks, and change archives; supports many coding agents.
|
|
- [Spec Kit](https://github.com/github/spec-kit) - GitHub's extensible toolkit for SDD, bug fixing, and idea assessment; provides a constitution-to-spec-to-plan-to-tasks workflow.
|
|
- [GSD Core](https://github.com/open-gsd/gsd-core) - Context-engineered SDD workflow with discuss, plan, execute, verify, and ship phases, including fresh-context subagents.
|
|
- [CCPM](https://github.com/automazeio/ccpm) - Agent skill that turns PRDs into epics, GitHub issues, and parallel worktree-based execution with traceability.
|
|
- [BMAD Method](https://github.com/bmad-code-org/BMAD-METHOD) - Role-based agent methodology spanning product discovery, planning, architecture, development, testing, and operations.
|
|
|
|
## Workflow And Methodology Systems
|
|
|
|
- [Superpowers](https://github.com/obra/superpowers) - Composable skills for brainstorming, detailed planning, TDD, review, debugging, worktrees, and subagent-driven development.
|
|
- [BigPowers](https://github.com/danielvm-git/bigpowers) - Prescriptive solo-developer lifecycle with specification state, TDD, quality gates, and an extensive cross-agent skill catalog.
|
|
- [Archon](https://github.com/coleam00/Archon) - Self-hosted YAML workflow engine for repeatable agent-powered development, approvals, checks, and audit trails.
|
|
- [Open GSD](https://github.com/open-gsd/gsd-core) - The active successor to the archived `gsd-build/get-shit-done` project; useful when evaluating GSD-style planning and execution.
|
|
- [Agent OS](https://github.com/buildermethods/agent-os) - Standards-first system that extracts repository conventions, injects relevant standards, and shapes implementation specifications across coding-agent tools.
|
|
- [MetaGPT](https://github.com/FoundationAgents/MetaGPT) - Role-based multi-agent workflow that turns a requirement into PRDs, designs, tasks, code, tests, and CI artifacts.
|
|
- [Ruflo](https://github.com/ruvnet/ruflo) - Agent meta-harness for Claude Code and Codex with reusable workflows and a SPARC-inspired phased development methodology with quality gates.
|
|
|
|
## Agent Skills By Purpose
|
|
|
|
### Discovery, Requirements, And Planning
|
|
|
|
- [Superpowers: brainstorming and writing-plans](https://github.com/obra/superpowers/tree/main/skills) - Elicit intent before implementation, then produce small, executable implementation tasks.
|
|
- [BigPowers](https://github.com/danielvm-git/bigpowers) - Includes context survey, research, specification elaboration, domain modeling, architecture, scope, and task-slicing skills.
|
|
- [CCPM](https://github.com/automazeio/ccpm) - Creates PRDs, technical epics, task dependencies, and GitHub issue hierarchies.
|
|
|
|
### Implementation, Testing, And Review
|
|
|
|
- [Superpowers](https://github.com/obra/superpowers) - Enforces red-green-refactor TDD, systematic debugging, verification before completion, and code-review workflows.
|
|
- [BigPowers](https://github.com/danielvm-git/bigpowers) - Provides `develop-tdd`, verification, audit, release, and engineering-convention skills within a gated lifecycle.
|
|
- [CCPM](https://github.com/automazeio/ccpm) - Coordinates independent implementation streams through Git worktrees and GitHub issues.
|
|
|
|
### Reverse Engineering And Existing-System Discovery
|
|
|
|
- [REA: Reverse Engineer Anything](https://github.com/morluto/rea) - MCP and skill-based local analysis of applications, websites, JavaScript/Electron bundles, binaries, APKs, and runtime behavior. Use only for software you own or are authorized to examine.
|
|
- [leanrea](https://github.com/aicodeking/leanrea) - A small, CLI-oriented skill for reverse-engineering workflows without an MCP server; a lightweight alternative to REA for agents with shell access.
|
|
|
|
### Mobile
|
|
|
|
#### Android
|
|
|
|
- [Awesome Android Agent Skills](https://github.com/new-silvermoon/awesome-android-agent-skills) - Curated skills for modern Android development with Kotlin, Jetpack Compose, architecture, testing, and Google Play release work.
|
|
- [Compose Performance Skills](https://github.com/skydoves/compose-performance-skills) - Focused Agent Skills library for diagnosing and improving Jetpack Compose performance.
|
|
- [Kotlin Multiplatform Agent Skills](https://github.com/mmiani/kotlin-kmp-claude-agent-skills) - Skills grounded in Kotlin Multiplatform, Compose, navigation, testing, and Android documentation.
|
|
|
|
#### iOS
|
|
|
|
- [iOS Agent Skills](https://github.com/rusel95/ios-agent-skills) - Evaluated skills for SwiftUI and UIKit architecture, Swift concurrency, testing, security, accessibility, localization, and logging.
|
|
- [Vercel Agent Skills: React Native Guidelines](https://github.com/vercel-labs/agent-skills/tree/main/skills/react-native-guidelines) - Cross-platform mobile guidance for React Native and Expo performance, layout, animation, images, state, and platform APIs.
|
|
|
|
### Game Development
|
|
|
|
- [Agentic Gamedev Skills](https://github.com/abagames/agentic-gamedev-skills) - Reusable skills extracted from game-development work and agentic workflow research.
|
|
- [Everything Game Dev Code](https://github.com/MRCalderon3D/everything-game-dev-code) - Multi-engine AI development scaffold with agents, commands, and skills for Unity, Unreal, Godot, and HTML games.
|
|
- [Defold Agent Config](https://github.com/indiesoftby/defold-agent-config) - `AGENTS.md` and skills for AI-assisted development with the Defold game engine.
|
|
- [Roblox Brain](https://github.com/TabooHarmony/roblox-brain) - Agent skill library for Roblox Studio development.
|
|
|
|
### Organization
|
|
|
|
#### Research
|
|
|
|
- [AI Research SKILLs](https://github.com/Orchestra-Research/AI-Research-SKILLs) - Broad open-source library of research and engineering skills for scientific workflows, literature work, experimentation, and technical writing.
|
|
- [Taskmaster](https://github.com/eyaltoledano/claude-task-master) - Supports project-aware research before task decomposition, allowing research findings to inform implementation tasks.
|
|
- [BigPowers](https://github.com/danielvm-git/bigpowers) - Includes research-first and context-survey skills for grounding plans in the existing codebase and external evidence.
|
|
|
|
#### Management
|
|
|
|
- [CCPM](https://github.com/automazeio/ccpm) - Project-management skill that connects PRDs, epics, GitHub issues, worktrees, parallel execution, and status reporting.
|
|
- [Beads](https://github.com/gastownhall/beads) - Dependency-aware task graph and persistent memory for coordinated long-horizon agent work.
|
|
- [Taskmaster](https://github.com/eyaltoledano/claude-task-master) - PRD-to-task management with dependencies, tags, research, and next-task workflows.
|
|
- [Archon](https://github.com/coleam00/Archon) - Workflow orchestration with approval gates, queues, audit records, and GitHub-integrated delivery.
|
|
|
|
#### Document Creation
|
|
|
|
- [Anthropic Document Skills](https://github.com/anthropics/skills/tree/main/skills) - Reference skills for creating and editing DOCX, PDF, PPTX, and XLSX files alongside technical and enterprise workflows.
|
|
- [Vercel Agent Skills: Writing Guidelines](https://github.com/vercel-labs/agent-skills/tree/main/skills/writing-guidelines) - Documentation writing and review guidance covering planning, tone, structure, code samples, and editorial consistency.
|
|
- [Overleaf Skill](https://github.com/aloth/overleaf-skill) - Agent skill for LaTeX project management, synchronization, and compilation with Overleaf.
|
|
|
|
### Skill Distribution
|
|
|
|
- [Agent Skills](https://agentskills.io) - Open specification for portable agent skills, including the `SKILL.md` format used by tools such as CCPM.
|
|
- [skills.sh](https://skills.sh) - Registry and CLI for discovering and installing skills across compatible coding agents.
|
|
- [Anthropic Skills](https://github.com/anthropics/skills) - Official examples of portable skills for development, testing, documents, design, and enterprise workflows.
|
|
- [Vercel Agent Skills](https://github.com/vercel-labs/agent-skills) - Official skills for React and Next.js performance, UI and accessibility reviews, React Native, documentation, deployment, and Vercel optimization.
|
|
- [Claude Code Workflows](https://github.com/wshobson/agents) - Multi-harness plugin marketplace with specialist agents, skills, and commands for development, testing, security, infrastructure, and review.
|
|
|
|
## Task, Context, And Knowledge Management
|
|
|
|
- [Beads](https://github.com/gastownhall/beads) - Distributed, dependency-aware issue tracker and persistent agent memory for long-running and multi-agent work.
|
|
- [Taskmaster](https://github.com/eyaltoledano/claude-task-master) - Turns PRDs into dependency-aware tasks, supports research and task expansion, and integrates through CLI and MCP.
|
|
- [CCPM](https://github.com/automazeio/ccpm) - GitHub-native project-management skill for tracing PRD, epic, issue, task, implementation, and progress.
|
|
- [GSD Core](https://github.com/open-gsd/gsd-core) - Preserves phase state across sessions while using fresh subagents to control context growth.
|
|
- [Aider](https://github.com/Aider-AI/aider) - Terminal pair-programming agent with repository maps, automatic Git commits, and configured lint/test feedback loops.
|
|
|
|
## Agent Runtimes And Orchestration
|
|
|
|
- [OpenCode](https://opencode.ai) - Open-source terminal coding agent with multi-provider support, skills, subagents, and MCP integration.
|
|
- [Claude Code](https://docs.anthropic.com/en/docs/claude-code/overview) - Anthropic's terminal coding agent with skills, plugins, hooks, subagents, and project instructions.
|
|
- [Codex CLI](https://github.com/openai/codex) - OpenAI's open-source coding agent for terminal workflows, sandboxing, skills, plugins, and MCP servers.
|
|
- [Gemini CLI](https://github.com/google-gemini/gemini-cli) - Google's open-source terminal agent with tool calling, extensions, and MCP support.
|
|
- [Cline](https://github.com/cline/cline) - Open-source VS Code coding agent with browser automation, MCP, and provider choice.
|
|
- [Roo Code](https://github.com/RooCodeInc/Roo-Code) - Open-source VS Code agent with specialized modes and MCP integration.
|
|
- [Cursor](https://cursor.com) - AI-first code editor with project rules, background agents, MCP, and a plugin marketplace.
|
|
- [Windsurf](https://windsurf.com) - Agentic IDE with repository rules, tool integration, and multi-step coding workflows.
|
|
- [Factory Droid](https://www.factory.ai) - Terminal and IDE coding-agent platform with skills, plugins, MCP, and parallel execution options.
|
|
- [OpenHands Agent Canvas](https://github.com/OpenHands/OpenHands) - Self-hosted control center for running local, Docker, remote, or cloud coding-agent backends and automations.
|
|
- [Archon](https://github.com/coleam00/Archon) - Self-hosted workflow engine for repeatable multi-step coding processes, human approval gates, and long-running queues.
|
|
|
|
## Verification, Quality, And Security
|
|
|
|
- [Playwright MCP](https://github.com/microsoft/playwright-mcp) - Accessibility-tree-based browser automation for agents to verify user flows, inspect console errors, and generate evidence.
|
|
- [Playwright CLI](https://github.com/microsoft/playwright-cli) - CLI and skills-oriented browser automation option for token-efficient coding-agent workflows.
|
|
- [Vercel Agent Skills](https://github.com/vercel-labs/agent-skills) - Includes React performance, web design, accessibility, UX, and documentation review skills.
|
|
- [Superpowers](https://github.com/obra/superpowers) - Supplies systematic debugging, TDD, verification-before-completion, and independent review workflows.
|
|
- [Semgrep](https://github.com/semgrep/semgrep) - Static analysis rules for code-quality and security gates in an implementation plan or CI workflow.
|
|
- [OSV-Scanner](https://github.com/google/osv-scanner) - Scans project dependencies and lockfiles for known vulnerabilities; useful as a reproducible acceptance check.
|
|
|
|
## Protocols And Repository Conventions
|
|
|
|
- [Model Context Protocol](https://modelcontextprotocol.io) - Open protocol for connecting agents to local and remote tools, data sources, and services.
|
|
- [Agent Skills](https://agentskills.io) - Portable `SKILL.md` packaging format for progressive, on-demand agent instructions.
|
|
- [AGENTS.md](https://agents.md) - Open convention for repository-local instructions that guide compatible coding agents.
|
|
- [Agent Client Protocol](https://agentclientprotocol.com) - Protocol for connecting coding-agent clients and agent servers; used by orchestration tools such as Agent Canvas.
|
|
- [MCP Registry](https://registry.modelcontextprotocol.io) - Official catalog for discovering MCP servers. Review server permissions and source code before installation.
|
|
|
|
## Platforms And Hosted Tools
|
|
|
|
- [Kiro](https://kiro.dev) - AWS's agentic IDE with requirements, design, and task artifacts built into its spec-driven workflow.
|
|
- [Tessl](https://tessl.io) - Specification-centric platform for AI-assisted software development and reusable code assets.
|
|
- [GitHub Copilot](https://github.com/features/copilot) - A coding-agent platform that can run repository-local instructions and integrate with Spec Kit workflows.
|
|
|
|
## Standards And Learning
|
|
|
|
- [OpenSpec documentation](https://openspec.dev/docs) - Guides for a lightweight, iterative SDD workflow and supported agent integrations.
|
|
- [Spec Kit: Spec-Driven Development](https://github.github.io/spec-kit/concepts/sdd.html) - Explanation of specification artifacts and the Spec Kit lifecycle.
|
|
- [Agent Skills specification](https://agentskills.io/specification) - Reference for authoring interoperable skills.
|
|
|
|
## Contributing
|
|
|
|
Suggestions should link to the canonical project page and state the tool's primary purpose, agent compatibility, license, and maintenance status. Prefer projects with public documentation, active maintenance, and artifacts that can be reviewed in version control.
|