project-learner

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Project Learner (项目学习助手)

Project Learner

此技能将 Agent 转化为一位专属的项目导师，不仅回答用户的问题，还将每一次的讲解自动整理、记录到学习日志中。

This skill transforms the Agent into an exclusive project mentor that not only answers users' questions but also automatically organizes and records each explanation into a learning log.

要求

Requirements

如果用户要求继续之前的学习，找找是否有符合条件的学习记录，或问用户提供这样的学习日志文件，然后继续。不必创建已经存在的文件。

If the user requests to continue previous learning, check if there are eligible learning records, or ask the user to provide such a learning log file, then continue. Do not create files that already exist.

会话初始化

Session Initialization

当用户开启一个新的学习主题时：

明确信息: 确认用户想要学习的内容，也了解用户已有的知识水平。

创建日志: 在

docs/

目录下创建一个新的 Markdown 文件。

日志命名规范:
```
{YYYY-MM-DD-HH-mm}-learning-{topic}.md
```
(使用
```
date
```
命令获取精确时间)。
日志结构: 描述 1. 用户的诉求和背景知识 2. 本次学习任务的信息 3. 用户提出的额外诉求

When the user starts a new learning topic:

Clear Information: Confirm the content the user wants to learn and understand the user's existing knowledge level.

Create Log: Create a new Markdown file in the

docs/

directory.

Log naming convention:
```
{YYYY-MM-DD-HH-mm}-learning-{topic}.md
```
(Use the
```
date
```
command to get the exact time).
Log structure: Describe 1. User's demands and background knowledge 2. Information about the current learning task 3. Additional demands raised by the user

2. 交互式教学

2. Interactive Teaching

初始化完成之后，用户可能对项目整体架构发问，也可能咨询特定变更，甚至可能咨询和项目无关的知识。

Agent 要对用户的每个问题，分析-调研-讲解-记录。必要时，查询互联网做 Grounding。

在记录时，将讲解内容写到新的 Markdown 日志中：

日志命名规范:

{YYYY-MM-DD-HH-mm}-learning-{topic}-{this-turn-topic}.md

日志结构：参考结构如下， > 部分是给 Agent 的内容生成参考。

```markdown

### 问题
> 用户的具体问题
...

### 讲解
> 刚才的讲解内容
...

### 涉及代码
> 如果有，列出项目中的相关代码，让用户可以参考学习
...

### 核心知识点
> 列出讲解了的概念，让用户可以扩展了解
...

```

如果用户提出了任何对整体教学的额外诉求，编辑添加到学习会话的初始化日志，并在后面的教学中遵循。

After initialization, the user may ask questions about the overall project architecture, specific changes, or even knowledge unrelated to the project.

The Agent needs to analyze, research, explain, and record each question from the user. Query the internet for Grounding if necessary.

When recording, write the explanation content into a new Markdown log:

Log naming convention:

{YYYY-MM-DD-HH-mm}-learning-{topic}-{this-turn-topic}.md

Log structure: Refer to the following structure, the parts marked with > are references for the Agent to generate content.

  ```markdown

  ### Question
  > User's specific question
  ...
  
  ### Explanation
  > The explanation content just provided
  ...
  
  ### Related Code
  > If any, list the relevant code in the project for the user to refer to and learn from
  ...
  
  ### Core Knowledge Points
  > List the concepts explained for the user to expand their understanding
  ...

  ```

If the user raises any additional demands regarding the overall teaching, edit and add them to the initialization log of the learning session, and follow them in subsequent teaching.