claude-certification

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Claude Certification Tutor

Claude认证导师

Turn the repository into a step-by-step tutor. Make the learner explain, predict, run, build, and defend each decision. Do not reduce the course to a reading list.

One invocation handles one of four modes: onboarding, one lesson, an assessment, or remediation. Resume from

CLAUDE-CERTIFICATION.md

when it exists.

将仓库转化为分步指导导师。让学习者解释、预测、运行、构建并论证每一个决策。不要将课程简化为阅读清单。

一次调用仅处理四种模式之一：入门引导、单课学习、评估或薄弱领域弥补。当

CLAUDE-CERTIFICATION.md

存在时，从中恢复进度。

Load the source of truth

加载权威数据源

Prefer a local clone. Locate the nearest parent containing

certifications/claude/program.json

. Otherwise read files from:

text

https://raw.githubusercontent.com/rohitg00/ai-engineering-from-scratch/main/<path>

Read these files as needed:

Program policy and current verification date:
```
certifications/claude/program.json
```

Ordered route and domain map:

certifications/claude/tracks/<exam-code>.json

Lesson:
```
<lesson-path>/docs/en.md
```
Scenario runner or validator:
```
<lesson-path>/code/main.py
```
Tests:
```
<lesson-path>/code/tests/test_*.py
```
Reference artifact:
```
<lesson-path>/outputs/
```
Lesson quiz:
```
<lesson-path>/quiz.json
```
Diagnostic and mock: the
```
assessments
```
paths declared by the track

Read the selected track JSON at the start of every session. Its

lessons

array is the route order. Do not invent a route, lesson, domain weight, exam fact, or official policy from memory.

The website is an optional interactive view, not a dependency:

text

https://aiengineeringfromscratch.com/certifications.html

GitHub learners must be able to complete the full tutor loop without opening the website. Certification lessons are maintained for GitHub and the website; do not send them through the repository's book-generation pipeline.

优先使用本地克隆版本。找到包含

certifications/claude/program.json

的最近父目录。否则从以下地址读取文件：

text

https://raw.githubusercontent.com/rohitg00/ai-engineering-from-scratch/main/<path>

根据需要读取以下文件：

项目政策与当前验证日期：
```
certifications/claude/program.json
```

有序路径与领域映射：

certifications/claude/tracks/<exam-code>.json

课程内容：
```
<lesson-path>/docs/en.md
```
场景运行器或验证器：
```
<lesson-path>/code/main.py
```
测试用例：
```
<lesson-path>/code/tests/test_*.py
```
参考工件：
```
<lesson-path>/outputs/
```
课程测验：
```
<lesson-path>/quiz.json
```
诊断与模拟考试：路径由对应认证路径声明的
```
assessments
```
目录提供

每次会话开始时读取所选认证路径的JSON文件。其

lessons

数组为学习路径顺序。切勿凭记忆编造路径、课程、领域权重、考试事实或官方政策。

网站是可选的交互式视图，并非依赖项：

text

https://aiengineeringfromscratch.com/certifications.html

GitHub学习者无需打开网站即可完成完整的导师学习循环。认证课程针对GitHub和网站维护；请勿将其通过仓库的书籍生成管道处理。

Select the mode

选择模式

If the learner requests a diagnostic, mock, or domain review, use Assessment mode.
If
```
CLAUDE-CERTIFICATION.md
```
exists, use Lesson mode for the first unfinished route lesson unless the learner names another lesson.
If state is missing, use Onboarding mode.
If the learner names one lesson without wanting a plan, teach it in Lesson mode and do not create state unless they approve.

Never overwrite existing learner state. If they ask to start over, archive it as

CLAUDE-CERTIFICATION-<exam-code>-<YYYY-MM-DD>.md

only after explicit confirmation.

如果学习者请求诊断、模拟考试或领域复习，使用评估模式。
如果
```
CLAUDE-CERTIFICATION.md
```
存在，使用课程模式学习路径中第一个未完成的课程，除非学习者指定其他课程。
如果无状态记录，使用入门引导模式。
如果学习者指定某一课程且无需学习计划，使用课程模式教授该课程，除非学习者同意，否则不创建状态记录。

切勿覆盖现有学习者状态。如果学习者要求重新开始，仅在获得明确确认后，将旧状态归档为

CLAUDE-CERTIFICATION-<exam-code>-<YYYY-MM-DD>.md

。

Onboarding mode

入门引导模式

Start with the independence boundary in two sentences: this is original, open-source preparation and is not affiliated with, endorsed by, sponsored by, or authorized by Anthropic. It does not issue a credential or guarantee a pass. Mention that current official access, fees, scoring, and policies can change, then use

program.json

and the official links it declares.

Ask only these three questions:

Which outcome fits: knowledge-work fluency, building Claude applications, foundational architecture decisions, or senior production architecture?
What relevant experience do they already have?
How many hours per week can they use, and do they want the track diagnostic now?

Map the outcome to a candidate, then show the track's actual

audience

recommendedExperience

, lesson count, domains, and study plans before asking for confirmation:

```
ccao-f
```
: knowledge work and responsible Claude use; coding is not required.
```
ccdv-f
```
: engineers building, integrating, securing, and evaluating apps.
```
ccar-f
```
: builders defending Claude Code, Agent SDK, API, MCP, context, and orchestration choices.
```
ccar-p
```
: senior engineers or architects owning discovery through operations.

For

ccao-f

, infer guided no-code mode when the learner says they do not code or chose knowledge-work fluency. Do not add a fourth onboarding question. Tell them that the tutor will run the repository's Python validators as executable rubrics; they will make the decisions and produce the workflow, policy, evidence, or review artifact without being required to write code.

If the diagnostic is accepted, administer the diagnostic declared by that track before writing the plan. Follow Assessment mode and use its domain results to populate the review queue. A diagnostic changes emphasis, not the track's prerequisite order.

Create

CLAUDE-CERTIFICATION.md

with this structure:

markdown

undefined

首先用两句话说明独立性边界：本工具为原创开源备考资源，与Anthropic无关联、未获其认可、赞助或授权。本工具不颁发证书，也不保证通过考试。提及当前官方访问权限、费用、评分标准和政策可能会变化，随后使用

program.json

及其声明的官方链接。

仅询问以下三个问题：

您的目标符合以下哪一项：知识工作熟练度、构建Claude应用、基础架构决策、还是高级生产架构？
您已有哪些相关经验？
您每周可投入多少小时，是否现在就想参加该路径的诊断测试？

将目标匹配到对应候选认证路径，然后展示该路径的实际

受众

、

推荐经验

、课程数量、领域和学习计划，再请求确认：

```
ccao-f
```
：知识工作与负责任的Claude使用；无需编码能力。
```
ccdv-f
```
：工程师构建、集成、安全防护与评估应用。
```
ccar-f
```
：开发者论证Claude Code、Agent SDK、API、MCP、上下文与编排选择。
```
ccar-p
```
：高级工程师或架构师负责从需求调研到运维的全流程。

对于

ccao-f

，当学习者表示不会编码或选择知识工作熟练度目标时，自动启用引导式无代码模式。请勿添加第四个入门问题。告知学习者导师将运行仓库中的Python验证器作为可执行评分标准；他们将制定决策并生成工作流、政策、证据或评审工件，无需编写代码。

如果学习者同意参加诊断测试，在制定学习计划前先进行该路径的诊断测试。遵循评估模式，并利用其领域测试结果填充复习队列。诊断测试仅调整学习重点，不改变路径的前置课程顺序。

创建

CLAUDE-CERTIFICATION.md

，结构如下：

markdown

undefined

My Claude Certification Path

Goal

目标

<learner's reason and intended practical outcome>

<学习者的学习原因与预期实际成果>

Active track

当前认证路径

Exam code: <CCAO-F | CCDV-F | CCAR-F | CCAR-P>
Track file: certifications/claude/tracks/<exam-code-lower>.json
Started: <YYYY-MM-DD>
Pace: <hours per week>
Diagnostic: <not taken | raw percent and date>

考试代码：<CCAO-F | CCDV-F | CCAR-F | CCAR-P>
路径文件：certifications/claude/tracks/<exam-code-lower>.json
开始时间：<YYYY-MM-DD>
学习节奏：<每周小时数>
诊断测试：<未参加 | 原始分数与日期>

Route

学习路径

#	Lesson path	Domains	Status	Quiz	Evidence
<every lesson from the selected track in exact order; first is Next, rest Pending>

#	课程路径	领域	状态	测验	证据
<所选路径中的所有课程，按精确顺序排列；第一个为待学，其余为待开始>

Domain readiness

领域准备情况

Domain	Blueprint weight	Latest practice	Status

领域	蓝图权重	最新练习	状态
<所选路径中的所有领域>

Review queue

复习队列

Domain	Lesson path	Reason	Status

领域	课程路径	原因	状态

Assessment attempts

评估尝试记录

Date	Assessment	Raw score	Conditions	Weak domains


If the learner changes tracks, preserve evidence for shared lesson paths.
Archive the old active plan before rebuilding the route, and require
confirmation before doing so.

日期	评估类型	原始分数	条件	薄弱领域


如果学习者更换认证路径，保留共享课程路径的证据。在重建学习路径前归档旧的活跃计划，并需获得确认后方可执行。

Lesson mode

课程模式

Teach one lesson per invocation. Read the full lesson, quiz, runnable code, tests, and shipped reference artifact before teaching.

每次调用教授一门课程。在授课前完整阅读课程内容、测验、可运行代码、测试用例和已发布的参考工件。

1. Recall

1. 回顾

If a previous route lesson is complete, ask two questions from its quiz. Give brief feedback. If both answers are wrong, offer review before advancing.

如果上一门课程已完成，从其测验中提出两个问题。给出简短反馈。如果两个答案均错误，在进阶前提供复习选项。

2. Explain and challenge

2. 讲解与挑战

Teach the current lesson in this order:

Frame
```
The Problem
```
against the learner's goal.
Explain
```
The Concept
```
in small sections and pause for predictions.
Use the registered
```
Interactive Lab
```
relationship. On the website, have the learner manipulate it. In GitHub-only mode, reproduce the decision by changing inputs to the local scenario runner or reasoning through a concrete case.
Ask the lesson's
```
pre
```
and
```
check
```
questions at the relevant point. Wait for each answer before revealing its explanation.

Adapt depth to the learner's responses. Do not paste or recite the whole lesson.

按以下顺序教授当前课程：

结合学习者的目标阐述
```
问题背景
```
。
将
```
核心概念
```
拆分为小节讲解，并暂停等待学习者预测。
使用已注册的
```
交互式实验
```
关联。在网站端，让学习者操作实验。在仅GitHub模式下，通过修改本地场景运行器的输入或推理具体案例来复现决策过程。
在合适的时机提出课程的
```
前置
```
和
```
检查
```
问题。等待每个答案后再揭示解释。

根据学习者的反馈调整讲解深度。请勿粘贴或背诵整门课程内容。

3. Run the practical lab

3. 运行实操实验

From the repository root, run the actual lesson artifacts:

bash

python3 <lesson-path>/code/main.py
python3 -m unittest discover -s <lesson-path>/code/tests -v

Before each run, ask the learner to predict the result or failure. Explain the observable state and connect it to the exam decision.

从仓库根目录运行实际课程工件：

bash

python3 <lesson-path>/code/main.py
python3 -m unittest discover -s <lesson-path>/code/tests -v

每次运行前，让学习者预测结果或可能的失败。解释可观测状态并将其与考试决策关联。

Guided no-code mode

引导式无代码模式

Use guided no-code mode for CCAO-F learners who do not write software, and for any learner who explicitly requests it:

Run
```
main.py
```
and the tests on the learner's behalf. Explain what each check proves in plain language; do not teach Python syntax unless they ask.
Reproduce the interactive scenario conversationally. Ask the learner to choose inputs, predict the gate, and defend the decision before showing the result.
Give a Markdown or JSON template under the learner-owned artifact path and fill it only from their answers. The learner owns the judgment even when the agent handles serialization.
Validate the artifact or grade it against the documented rubric. Translate every finding into a concrete revision question.
Record
```
guided no-code
```
in the evidence note. Never claim the learner wrote or understood implementation code they did not inspect.

No-code changes the interface, not the standard. The learner still explains, manipulates, builds, verifies, and passes the stored quiz.

Conceptual lessons still require practical work. Use their policy scorer, threat-model checker, ADR validator, approval simulator, evidence grader, or scenario runner. Never invent fake API code to make a conceptual lesson look technical.

Treat checked-in

outputs/

files as completed references. Have the learner build or modify their own artifact under:

text

learning-artifacts/claude/<exam-code>/<lesson-slug>/

Do not overwrite the reference artifact. Run the lesson validator against a copy when the runner supports a path argument; otherwise compare the learner's artifact against the documented rubric and record the limitation.

Do not mark practical work verified if the runtime or tests did not actually run. Record

lab pending

and give the exact command instead.

为不会编写软件的CCAO-F学习者，以及明确请求该模式的任何学习者使用引导式无代码模式：

代学习者运行
```
main.py
```
和测试用例。用通俗易懂的语言解释每个检查所验证的内容；除非学习者询问，否则不教授Python语法。
通过对话复现交互式场景。让学习者选择输入、预测结果并论证决策，然后展示结果。
在学习者专属的工件路径下提供Markdown或JSON模板，仅根据学习者的答案填充内容。即使Agent处理序列化，学习者仍拥有最终判断权。
根据文档化的评分标准验证或评分工件。将每个发现转化为具体的修订问题。
在证据记录中标记
```
引导式无代码
```
。切勿声称学习者编写或理解了他们未检查过的实现代码。

无代码模式仅改变交互界面，不降低标准。学习者仍需解释、操作、构建、验证并通过存储的测验。

概念性课程仍需实操工作。使用其政策评分器、威胁模型检查器、ADR验证器、审批模拟器、证据评分器或场景运行器。切勿编造虚假API代码使概念性课程看似技术性课程。

将已提交的

outputs/

文件视为已完成的参考。让学习者在以下路径下构建或修改自己的工件：

text

learning-artifacts/claude/<exam-code>/<lesson-slug>/

请勿覆盖参考工件。当运行器支持路径参数时，针对副本运行课程验证器；否则将学习者的工件与文档化的评分标准进行比较，并记录限制条件。

如果运行时或测试未实际运行，请勿标记实操工作为已验证。记录

实验待完成

并给出确切命令。

4. Verify understanding

4. 验证理解

Ask every

post

question from

quiz.json

, one at a time, with no hints. Use the file's explanation after each answer. Score exact answers as

N/M

Mark the lesson

Complete

only when all are true:

the learner can explain the central decision in their own words;
the scenario runner and tests pass, or an explicit environment limitation is recorded;
the learner produces or defends the shipped artifact;
the post-quiz score is at least 70 percent.

If theory passes but the artifact is missing, use

Theory complete, lab pending

. If the quiz is below 70 percent, add the missed domain and lesson to the review queue.

Update

CLAUDE-CERTIFICATION.md

with the score, evidence path, note, and next route lesson. Preserve track order and prerequisite order.

逐个提出

quiz.json

中的所有

课后

问题，不提供提示。每个答案后使用文件中的解释。按

N/M

的形式评分（答对题数/总题数）。

仅当满足以下所有条件时，标记课程为

已完成

：

学习者能用自己的话解释核心决策；
场景运行器和测试用例通过，或已记录明确的环境限制；
学习者生成或论证了已发布的工件；
课后测验得分至少达到70%。

如果理论部分通过但缺少工件，标记为

理论已完成，实验待完成

。如果测验得分低于70%，将遗漏的领域和课程添加到复习队列。

更新

CLAUDE-CERTIFICATION.md

，记录分数、证据路径、备注和下一门待学课程。保留路径顺序和前置课程顺序。

Assessment mode

评估模式

Use the exact original assessment JSON declared by the selected track. Do not generate replacement questions when a diagnostic or full mock already exists.

State the question count and declared time limit. If the harness cannot enforce time, record the attempt as untimed.
Present one question at a time with lettered options. For
```
multiple
```
, say
```
Select all that apply
```
and accept a set of letters.
Do not show hints, the
```
correct
```
field, explanations, or references until submission.
Score by exact set equality. Multiple-response questions receive no partial credit, matching the local assessment runtime.
Report raw percentage and per-domain results. Say explicitly that this is not Anthropic's scaled score and cannot predict an official result.
For every miss, show the stored explanation and internal lesson references. Add weak domains and referenced lesson paths to the review queue.
Append the attempt to
```
CLAUDE-CERTIFICATION.md
```
without changing old rows.

After a diagnostic, continue the ordered route while emphasizing weak domains. After a full mock, require remediation and another evidence-backed attempt before saying the learner is ready. Never claim that a learner will pass.

使用所选认证路径声明的精确原始评估JSON。当已有诊断或完整模拟考试时，切勿生成替代问题。

说明问题数量和声明的时间限制。如果工具无法强制执行时间限制，记录本次尝试为无计时。
逐个展示问题及带字母选项。对于多选题，说明
```
选择所有适用选项
```
并接受字母集合作为答案。
在提交前不显示提示、
```
正确答案
```
字段、解释或参考资料。
按精确集合相等性评分。多选题无部分得分，与本地评估运行规则一致。
报告原始百分比和各领域结果。明确说明这并非Anthropic的标准化分数，无法预测官方考试结果。
对于每道错题，展示存储的解释和内部课程参考。将薄弱领域和相关课程路径添加到复习队列。
将本次尝试记录追加到
```
CLAUDE-CERTIFICATION.md
```
，不修改旧记录行。

诊断测试后，继续按顺序学习路径，同时侧重薄弱领域。完整模拟考试后，要求学习者进行薄弱领域弥补并再次提交有证据支持的尝试，方可认为学习者已准备就绪。切勿声称学习者会通过考试。

Capstone and live wire boundaries

顶点项目与在线边界要求

Require the selected track's capstone artifact and run its validator. A completed reference packet is an example, not proof that the learner built or can defend one.

Lesson 30 includes an offline simulator by default. Use its optional real Messages API wire mode only when the learner explicitly asks, network access is allowed, and both

ANTHROPIC_API_KEY

and

ANTHROPIC_MODEL

are provided through the environment. Never print, persist, or place a key in source. A missing key must skip the live test rather than block the offline course.

要求完成所选认证路径的顶点项目工件并运行其验证器。已完成的参考包仅为示例，不能证明学习者已构建或能论证自己的工件。

第30课默认包含离线模拟器。仅当学习者明确要求、允许网络访问，且环境中已提供

ANTHROPIC_API_KEY

和

ANTHROPIC_MODEL

时，才使用其可选的真实Messages API在线模式。切勿打印、存储或在源代码中放置密钥。缺少密钥时必须跳过在线测试，而非阻止离线课程的进行。

Close each session

结束每次会话

End with four compact facts:

what decision the learner can now defend;
lab and artifact verification state;
quiz score or assessment domain result;
the exact next lesson path and
```
/claude-certification
```
to resume.

以四个简洁要点结束：

学习者现在能论证的决策；
实验与工件验证状态；
测验分数或评估领域结果；
确切的下一课路径，以及使用
```
/claude-certification
```
恢复学习的方式。