DevCrew_s is the specification repository for the DevCrew multi-agent framework. This repository contains all agent specifications, protocol definitions, and architectural documentation that define how autonomous AI agents collaborate to perform complex development tasks across various domains.
Note: This is a specification-only repository. For actual implementations, see the DevCrew_i repository.
AI agents augment, not replace, human expertise. Our framework embodies the principle that AI and humans excel in different areas:
- Humans excel at: Creativity, ethical judgment, strategic vision, and emotional intelligence (EPOCH Framework)
- AI agents excel at: Pattern recognition, repetitive tasks, 24/7 availability, and processing vast amounts of data
Research shows AI agents can boost team productivity by 60% while humans maintain control over strategic decisions. Learn more about AI augmentation vs. replacement →
- 79% of companies are already adopting AI agents
- 96% of organizations plan to expand AI usage in 2025
- Early adopters achieve 30-50% cost reductions while improving quality
Explore the full business case →
- Specialized Roles: Each agent has focused expertise
- Protocol-Driven: Standardized workflows ensure consistency
- Human Oversight: Strategic decisions remain with humans
- Reusable: One protocol serves multiple agents
- Composable: Protocols can invoke other protocols
- Auditable: Clear success criteria and failure handling
- Quality Gates: Automated validation checkpoints
- Architecture Fitness: Continuous compliance verification
- Privacy by Design: Built-in data protection
| Agent | Role | Key Protocols |
|---|---|---|
| Backend-Engineer | Server-side development | GH-1, P-TDD, P-RECOVERY |
| Code-Reviewer | Quality assurance | Code Review, Security Scan, Feedback Generation |
| Product-Owner | Product strategy | RICE Scoring, User Story Mapping, Roadmap Sync |
| QA-Tester | Testing automation | P-TDD, Test Generation, Validation |
| System-Architect | Architecture governance | ADR Creation, ASR Extraction, Technology Radar |
DevCrew agents collaborate through 48 standardized protocols across 8 categories:
- P-TDD: Test-Driven Development enforcement
- GH-1: GitHub Issue Triage workflow
- P-RECOVERY: Failure recovery and rollback
- ADR Creation: Architecture decision records
- P-DEVSECOPS: Integrated security pipeline
DevCrew specifications can be implemented using various platforms:
- AWS Strands - Enterprise-scale AWS implementation
- Claude Code - Anthropic's AI coding assistant
- CrewAI - Open-source multi-agent framework
-
Clone the specifications repository
git clone <repository-url> cd devCrew_s
-
Study the core agent specifications
- agent-Backend-Engineer-vSEP25.md - Development workflows
- agent-System-Architect-vSEP25.md - Architecture governance
- agent-QA-Tester-vSEP25.md - Test automation
-
Understand the protocol structure
- Begin with GH-1 GitHub Issue Triage
- Study P-TDD Test-Driven Development
- Review all available protocols
-
Choose your implementation approach
- Review the implementation guides for your platform
- Fork the DevCrew_i repository for implementation
- Follow platform-specific setup instructions
-
Contribute improvements
- Propose new agents via agent proposals
- Submit protocol enhancements
- Share implementation experiences
Your contributions make DevCrew better for everyone! Whether you're fixing a typo, proposing a new protocol, or designing an innovative agent, every contribution matters.
- Shape the Future: Help define how AI agents collaborate in development workflows
- Learn & Grow: Gain expertise in AI agent architecture and protocols
- Community Impact: Your improvements benefit thousands of developers
- Recognition: Contributors are credited in our releases and documentation
- 🐛 Report Issues: Found a bug or inconsistency? Open an issue
- 📝 Fix Documentation: Improve clarity, fix typos, add examples
- 💡 Suggest Ideas: Propose enhancements via issues
- 🔄 Improve Protocols: Enhance error handling, add edge cases
- 🤖 Review Agent Specs: Validate proposed agents, suggest improvements
- 📚 Write Guides: Share implementation experiences and best practices
- 🆕 Create New Protocols: Design reusable workflows for common tasks
- 🤖 Propose New Agents: Use our AI Agent Specification Template
- 🏗️ Architectural Improvements: Propose system-level enhancements
-
AI Agent Adoption Guide - Comprehensive business case
-
Protocol Registry - Complete protocol documentation
-
Implementation Guides - Platform-specific tutorials
-
SECURITY.md - Security policy and vulnerability reporting
-
📖 Documentation: Start with our implementation guides
-
🐛 Bug Reports: Report issues
-
💡 Feature Requests: Suggest improvements
-
📧 Contact: Reach out through issue tracker or pull requests
This project is in the worldwide public domain. As stated in CONTRIBUTING:
This project is in the public domain within the United States, and copyright and related rights in the work worldwide are waived through the CC0 1.0 Universal public domain dedication.
All contributions to this project will be released under the CC0 dedication. By submitting a pull request, you are agreeing to comply with this waiver of copyright interest.
Ready to contribute? Start with a good first issue or create an issue with your ideas. Whether you're from software, product, operations, or any other domain, your expertise can help shape the future of AI-assisted development.