30-day CCA-F study plan

• What is a Large Language Model? (F)

• Meet Claude — the API model (F)

• Prompt engineering, the basics (F)

• What is a tool? (Function calling) (F)

• Agents 101 — LLMs in a loop (F)

• MCP — Model Context Protocol intro (F)

Foundations complete.

• 1.1 · Agent, workflow, or conversation? (D1)

• 1.2 · Agentic Loops & stop_reason (D1)

• 1.3 · Orchestration Patterns (D1)

• 1.4 · Multi-Agent Systems & Subagents (D1)

• 1.5 · Guardrails, hooks, and safe termination (D1)

• 1.6 · Failure modes & debugging agents (D1)

D1 lessons complete — the highest-weight domain (27%).

D1 sprint quiz →

D1 sprint quiz, then re-read exam traps on any lesson you missed questions from.

• 2.1 · Claude Code architecture (D2)

• 2.2 · CLAUDE.md — hierarchy & precedence (D2)

• 2.3 · Skills, slash commands, subagents (D2)

• 2.4 · Hooks — lifecycle & exit codes (D2)

• 2.5 · SDK & CI/CD integration (D2)

D2 lessons complete.

D2 sprint quiz →

D2 sprint quiz.

• 3.1 · Prompt fundamentals (D3)

• 3.2 · Few-shot, chain-of-thought, extended thinking (D3)

• 3.3 · Structured output via tool use (D3)

• 3.4 · Evaluation — how you know it works (D3)

• 3.5 · Prompt caching for cost & latency (D3)

D3 lessons complete.

D3 sprint quiz →

D3 sprint quiz.

• 4.1 · Tool design fundamentals (D4)
• 4.2 · MCP architecture (D4)

• 4.3 · MCP in practice (D4)

D4 sprint quiz →

D4 lessons complete, then D4 sprint quiz.

• 5.1 · Context window fundamentals (D5)
• 5.2 · RAG and retrieval patterns (D5)

• 5.3 · Multi-instance & multi-pass review (D5)
• 5.4 · Cost, latency, and reliability (D5)

D5 sprint quiz →

D5 lessons complete, then D5 sprint quiz — all 29 lessons done.

Full mock exam →

Full mock exam (60 Q, 120 min). Review every missed question and re-read that lesson before the real exam.