New: The Agentic SDLC Handbook — the in-depth reference for PROSE. Free, online. Read it →

Reference

📖 Deep dive in the handbook: The Agentic SDLC Handbook → This page is the quick reference: checklists, links, and troubleshooting. For methodology, governance, and the full theoretical grounding, the handbook is the source.

Essential checklists, progression frameworks, and documentation links for ongoing AI Native Development implementation and mastery.

Quick Start Checklist

Conceptual Foundation

  1. [ ] Understand Markdown Prompt Engineering principles (semantic structure + precision + tools)
  2. [ ] Grasp Context Engineering fundamentals (buffer optimization + session strategy)
  3. [ ] Learn Skills as Distribution (capabilities packaged for auto-discovery)

Implementation Steps

  1. [ ] Install useful Skills for your stack: apm install owner/skill-name (Skills: instant capabilities)
  2. [ ] Create .github/copilot-instructions.md with basic project guidelines (Context Engineering: global rules)
  3. [ ] Set up domain-specific .instructions.md files with applyTo patterns (Context Engineering: selective loading)
  4. [ ] Configure Custom Agents (.agent.md) for your tech stack domains (Context Engineering: domain boundaries)
  5. [ ] Create first .prompt.md file with validation checkpoints (Markdown Prompt Engineering: deterministic templates)
  6. [ ] Build your first .spec.md template for feature specifications (Agent Primitive: deterministic planning-to-implementation bridge)
  7. [ ] Package reusable patterns as a Skill: apm init skill (Skills: share knowledge)
  8. [ ] Practice spec-first workflow: plan first, implement second (Context Engineering: session splitting)
  9. [ ] Test async delegation with GitHub Coding Agent (Advanced orchestration)
  10. [ ] Establish team governance and validation gates (Human-AI collaboration patterns)

Mastery Progression

Foundation Level

Goal: Understand core concepts and build first primitives
Time Investment: 2-3 hours
Key Outcomes:

  • Created basic .instructions.md files
  • Configured first Custom Agent
  • Built first .prompt.md template
  • Understand theoretical framework

Next Step: Move to Beginner level with hands-on implementation

Beginner Level

Goal: Basic instructions and prompts working consistently
Time Investment: 4-6 hours
Key Outcomes:

  • Working domain-specific instructions
  • Multiple Custom Agents configured
  • Prompt library with 3-5 templates
  • Consistent AI interactions in daily work

Next Step: Advance to Intermediate with workflow patterns

Intermediate Level

Goal: Spec-driven workflows with context optimization
Time Investment: 8-12 hours
Key Outcomes:

  • Spec-first planning methodology
  • Context engineering strategies implemented
  • Session splitting for complex tasks
  • Memory-driven development patterns

Next Step: Progress to Advanced with async delegation

Advanced Level

Goal: Async delegation and multi-agent orchestration
Time Investment: 15-20 hours
Key Outcomes:

  • GitHub Coding Agent delegation working
  • Parallel multi-agent workflows
  • Quality gates and validation processes
  • Hybrid context strategies

Next Step: Achieve Expert level with team implementation

Expert Level

Goal: Team-wide governance and frontier pattern innovation
Time Investment: 25+ hours
Key Outcomes:

  • Team-scale coordination frameworks
  • Knowledge sharing systems
  • Governance and compliance integration
  • Innovation in AI Native Development patterns

Next Step: Contribute to community knowledge and frontier research

The Paradigm Shift

Traditional approach: “Tell the AI what to do”
Agent Mastery approach: “Engineer the context and structure for optimal cognitive performance”

Core Principles

  1. Determinism through Structure: More predictable outcomes through systematic approaches
  2. Context as Performance: Strategic memory management for optimal AI cognitive performance
  3. Compound Intelligence: Systems that improve through iteration and learning
  4. Human-AI Partnership: Validation gates and collaborative workflows
  5. Team-Scale Coordination: Knowledge sharing and organizational transformation

Key Insights

  • The more determinism you need, the more Markdown Prompt Engineering and smaller scope you must use
  • The more complex your project, the more Context Engineering becomes critical
  • Master both principles and you’ll achieve unprecedented consistency and quality in agent-driven development

Remember: Start simple, iterate fast, scale systematically through systematic application of these frontier concepts.

Documentation References

Agent Skills & Distribution

Community Resources

  • Awesome GitHub Copilot - Comprehensive catalog of community-contributed instructions, prompts, and Custom Agents across all major languages and frameworks

VSCode Copilot Customization

GitHub Copilot Documentation

Quick Troubleshooting

Common Issues & Solutions

Issue: AI responses are inconsistent
Solution: Implement more structured Markdown prompts with clear headers and validation gates

Issue: Context window limitations
Solution: Use session splitting and modular instructions with applyTo patterns

Issue: Team coordination conflicts
Solution: Establish shared primitive libraries and repository coordination protocols

Issue: Quality concerns with async agents
Solution: Implement validation gates and treat agent outputs as high-quality drafts requiring review

Issue: Compliance and security concerns
Solution: Use risk-based agent boundaries and embed policy requirements in instructions

Success Metrics Tracking

Individual Metrics

  • Productivity: Time saved per feature implementation
  • Quality: Reduction in bugs and rework cycles
  • Consistency: Standardization of AI interaction patterns
  • Learning: Speed of new technique adoption

Team Metrics

  • Coordination: Reduction in merge conflicts and coordination overhead
  • Knowledge Sharing: Primitive reuse rates across team members
  • Onboarding: Time to productivity for new team members
  • Innovation: Development of new AI Native patterns

Organizational Metrics

  • Adoption: Percentage of teams using AI Native Development
  • Compliance: Adherence to governance frameworks
  • ROI: Overall productivity and quality improvements
  • Innovation: Contribution to frontier AI development practices

Next Steps

Immediate Actions (This Week)

  1. Complete the Quick Start Checklist
  2. Choose your learning path from the main guide
  3. Implement your first Agent Primitive

Short Term (This Month)

  1. Build a complete primitive library for your domain
  2. Practice workflow orchestration patterns
  3. Share knowledge with team members

Medium Term (Next Quarter)

  1. Implement team-scale coordination
  2. Establish governance frameworks
  3. Measure and optimize success metrics

Long Term (This Year)

  1. Achieve expert-level mastery
  2. Contribute to community knowledge
  3. Drive organizational AI Native transformation

This reference guide provides the essential resources for your AI Native Development journey. Return here whenever you need quick lookups, progress tracking, or implementation guidance.