In this Claude Certified Developer Foundations (CCDV-F) Certification Prep course, you'll learn:
Domain 1: Agents and Workflows
- Choose between a workflow and an agent based on task requirements
- Design manager/supervisor hierarchies and apply subagents
- Build agents using the Claude Agent SDK and custom agent loops
- Select an agent deployment model and apply hooks for deterministic actions
- Apply agent design patterns, including tool-use loops, memory, and context management
- Evaluate agentic frameworks (e.g., LangGraph, PydanticAI) for multi-step workflows
Domain 2: Applications and Integration
- Translate business requirements into functional and infrastructure requirements
- Apply systems life cycle concepts to develop and maintain Claude-based systems
- Apply Claude API mechanics, including messages, tools, streaming, and caching
- Select between realtime and batch API access patterns
- Apply core software engineering practices, including REST APIs, JSON, and version control
- Integrate Claude development into the SDLC and refactoring workflows
- Design applications that account for how Claude interprets instructions across interfaces
- Apply content boundaries, schema design, and session hygiene
- Configure Claude components using CLAUDE.md, settings.json, and version pinning
Domain 3: Claude Code
- Operate Claude Code components, including Rules, Skills, Commands, and Agents
- Manage Claude Code sessions using slash commands, headless mode, and auto-mode
- Configure the CLAUDE.md hierarchy and settings.json for a repository
Domain 4: Eval, Testing, and Debugging
- Identify error types and select an appropriate recovery strategy
- Analyze traces to identify failure modes
- Isolate problems between the integration layer and model output
Domain 5: Model Selection and Optimization
- Explain foundational LLM concepts: tokens, context windows, sampling, and non-determinism
- Select model options such as extended thinking and effort levels
- Apply fundamental prompting techniques: zero-shot, single-shot, and multi-shot
- Integrate Claude using SDKs that wrap REST APIs and websockets
- Differentiate use cases for Opus, Sonnet, and Haiku
- Weigh quality, latency, and cost tradeoffs when selecting a model
- Track token usage and apply caching techniques to optimize cost
Domain 6: Prompt and Context Engineering
- Manage the context window and prevent context drift and bloat
- Apply context isolation using subagents or multi-step workflows
- Apply prompt engineering principles, including few-shot examples and output constraints
- Iteratively refine and sanitize prompts and inputs
- Apply structured output patterns and validate Claude's responses
- Apply defensive parsing and skepticism toward confident output
Domain 7: Security and Safety
- Apply security best practices, including prompt injection and jailbreak defense
- Prevent data leakage and apply appropriate PII handling
- Layer guardrails using secure-by-design principles, including least privilege
- Use hooks to prevent destructive actions
- Manage secrets, credentials, and API keys across environments
- Apply identity validation and authorized-access monitoring
Domain 8: Tools and MCPs
- Implement tool use and function calling for external systems
- Apply tool usage patterns, including client-side vs. server-side and approval patterns
- Apply tool set construction and error-handling best practices
- Author and deploy MCP servers, including tools, resources, and prompts
- Apply MCP communication patterns, including stdio and client-server design
- Weigh tradeoffs among built-in tools, custom tools, Skills, and MCPs
- Â Â
