基于场景的考试形式与评分
选择题场景，达到 720/1000 分即可通过，不因猜错扣分。
基于场景的考试形式与评分 是 CoddyKit 上的免费 Claude Architect 课时。 这是第 3 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 Claude Architect 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Claude Architect 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
What the Exam Looks Like
The exam is scenario-based multiple choice: read a realistic situation, then pick the best of 4 options where exactly 1 is correct. No code to write, no essays.
How Scoring Works
Your result is a scaled score from 100 to 1000, with the pass mark at 720. Scaling keeps different exam forms comparable — just aim well above 720 for safety.
# Scaled score range, not a raw percentage
MIN_SCORE = 100
MAX_SCORE = 1000
PASS_SCORE = 720
my_score = 760
print("PASS" if my_score >= PASS_SCORE else "FAIL")4 of 8 Scenarios
The exam draws from 8 core scenarios, but any single sitting shows you 4 of them — and you won't know which in advance. The takeaway: study all 8.
The 8 Scenarios at a Glance
The 8 scenarios each map to a real pattern: customer support, code generation, multi-agent research, developer productivity, CI/CD, data extraction, conversational AI, and agentic tools.
No Penalty for Guessing
There's no penalty for guessing — a wrong answer costs the same as a blank: zero. So never leave one empty. Eliminate what you can, then pick the best of the rest.
# No guessing penalty -> always answer
# Expected value of guessing 1-of-4 is strictly positive
blank_value = 0
guess_value = 1 / 4 # > 0, so guessing beats leaving blank
print("Answer every question:", guess_value > blank_value)Who the Exam Is For
The audience is solution architects with 6+ months of production Claude experience. That's why questions are scenario-based — they test the judgment that comes from shipping, not memorized facts.
The 5 Domains and Their Weights
Questions spread across 5 weighted domains: D1 27%, D2 18%, D3 20%, D4 20%, D5 15%. D1 carries the most weight, so agent design and orchestration deserve the deepest study.
DOMAIN_WEIGHTS = {
"D1 Agent Architecture & Orchestration": 27,
"D2 Tool Design & MCP": 18,
"D3 Claude Code Config & Workflows": 20,
"D4 Prompt Engineering & Structured Output": 20,
"D5 Context Management & Reliability": 15,
}
print(sum(DOMAIN_WEIGHTS.values())) # 100Why Scenarios Beat Trivia
A trivia question asks what stop_reason returns; a scenario drops you mid-loop and asks what to do next. You're given a real situation and must choose the architecturally correct response.
Distractors Are Often Anti-Patterns
Wrong options are written to be tempting — many are classic anti-patterns: parsing text for 'done', iteration caps as the main stop, or enforcing rules with prompts not hooks. Spot one, rule it out.
Read for the Strongest Answer
Often more than one option is technically valid; the exam wants the strongest for the context. Anchor on principles: terminate on stop_reason, use hooks for critical rules, pass context to subagents.
# Correct agentic-loop termination: driven by stop_reason
while True:
resp = client.messages.create(model=MODEL, max_tokens=1024,
messages=history, tools=tools)
if resp.stop_reason == "tool_use":
history.append({"role": "assistant", "content": resp.content})
history.append(run_tools(resp)) # append results, continue
continue
break # end_turn / max_tokens / stop_sequence -> stop hereA Smart Test-Taking Strategy
A smart strategy: answer everything, eliminate anti-patterns to shrink 4 options to 2, weight prep toward D1, study all 8 scenarios, and aim past 720 with margin.
Quick Check: Format & Scoring
Apply what you've learned about the exam's format and scoring rules.
Recap: Format & Scoring
Recap: scenario-based MCQ (4 options, 1 correct), scaled 100-1000, pass 720, 4 of 8 scenarios, no guessing penalty. Domains D1 27% to D5 15%. Eliminate anti-patterns, then pick the strongest.
用 AI 导师学习 Python — 免费
在浏览器中编写并运行真实代码，获得全天候 AI 导师的即时帮助，并在网页或应用中继续学习。
- 课程
- 26
- 课程
- 104
常见问题解答
「基于场景的考试形式与评分」课时是免费的吗？
是的 — 「基于场景的考试形式与评分」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Claude Architect 课程的其余内容，请升级到 CoddyKit PRO。 Claude Architect 课程共包含 4 节课。
「基于场景的考试形式与评分」这节课中我会学到什么？
选择题场景，达到 720/1000 分即可通过，不因猜错扣分。 你通过在浏览器中直接运行的动手代码来练习 Claude Architect，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 Claude Architect 需要有经验吗？
无需任何先前经验。CoddyKit 上的 Claude Architect 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 3 节课，共 4 节。
「基于场景的考试形式与评分」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 Claude Architect 课中编写并运行代码吗？
能。每节 Claude Architect 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。