CCDV-F cheat sheet
The fields, values and orderings worth holding in memory the morning of the exam.
Exam facts
| Scored questions | 53 | Time | 120 minutes |
|---|---|---|---|
| Pass mark | 720 / 1,000 scaled | Format | Multiple choice and multiple response |
| Biggest domain | Applications and Integration, 33.1% | Smallest | Eval and Debugging, 2.6% |
Domain weights, in order
| Weight | Domain |
|---|---|
| 33.1% | Applications and Integration |
| 16.8% | Model Selection and Optimisation |
| 14.7% | Agents and Workflows |
| 11.0% | Prompt and Context Engineering |
| 10.6% | Tools and MCPs |
| 8.1% | Security and Safety |
| 3.1% | Claude Code |
| 2.6% | Eval, Testing and Debugging |
Stop reasons
| Value | Do |
|---|---|
end_turn | Use the response |
tool_use | Execute, append tool_result, loop |
max_tokens | Truncated — do not parse as final |
stop_sequence | Hit your stop string |
refusal | Safety stop; may not match your schema |
pause_turn | Continue the turn |
model_context_window_exceeded | Trim input |
Retry matrix
| Retry | 429 (honour retry-after), 500, 504, 529 |
|---|---|
| Do not retry | 400, 401, 403, 404, 413 |
| Size caps | 32 MB Messages · 256 MB Batch · 500 MB Files |
The five workflow patterns
| Pattern | Use when |
|---|---|
| Prompt chaining | Fixed, decomposable steps; gate between them |
| Routing | Distinct input categories, or cost-based model selection |
| Parallelisation — sectioning | Independent subtasks run at once |
| Parallelisation — voting | Same task repeated for confidence |
| Orchestrator–workers | Subtasks not known in advance |
| Evaluator–optimiser | Clear criteria and iteration measurably helps |
Numbers to know
| Cache breakpoints | 4 max · order tools → system → messages |
|---|---|
| Cache pricing | read 0.1× · write 1.25× (5 min) or 2× (1 h) |
| ITPM | Cache reads do not count toward it |
| Batches | 50% off · ≤100,000 requests / 256 MB · ≤24 h · no streaming |
| Strict tools | ≤20 per request · ≤24 optional params · ≤16 using anyOf |
| Tool overhead | +286 tokens (auto/none) · +406 (any/tool) |
| Thinking budget | Minimum 1,024 tokens; must leave room under max_tokens |
| Stop sequences | Maximum 5 |
| Context editing defaults | Trigger 100,000 input tokens · keep 3 tool uses |
Schema support
| Supported | types, properties, required, additionalProperties, enum, const, anyOf, oneOf, type arrays, items |
|---|---|
| Not supported | pattern, allOf, not, if/then/else, dependentSchemas, recursion |
MCP quick reference
| Server primitives | Tools · Resources · Prompts |
|---|---|
| Client primitives | Sampling · Elicitation · Logging |
| Transports | stdio (local, one client) · Streamable HTTP (remote, many) |
| Tool naming | mcp__<server>__<tool> |
| Shared config | .mcp.json at the repo root, with ${VAR} expansion |
| stdio rule | Never write to stdout |
← Previous: The 13 anti-patterns | Next: Questions & answers — Domain 1 →
Explore Other Certification Study Guides
Switch to another exam guide to prepare for your next Claude certification:
Associate
CCAO-F
Claude Certified Associate — Foundations
Entry-level certification for non-coders: prompting, output evaluation, model selection, workflow integration, configuration, governance & responsible use.
Open Guide →
7 Domains
Architect
CCAR-F
Claude Certified Architect — Foundations
Core architectural principles: agentic loops & orchestration, tool design & MCP, Claude Code workflows, prompt engineering, and context reliability.
Open Guide →
5 Domains
Architect
CCAR-P
Claude Certified Architect — Professional
Advanced enterprise architecture: solution design, system integration, evaluation frameworks, governance, safety & developer enablement.
Open Guide →
7 Domains