fix(chapter7): align reflection test task with code-generation prompts

[qicaiy]

Summary

• update code/chapter7/test_reflection_agent.py example to use code_agent for the final test run
• replace article-writing prompt with a code-generation task so it matches the custom code prompts

Why

Issue #362 points out that the example defines code-oriented prompts but runs a writing task with general_agent, which is confusing for learners.

Fixes #362

[Nikolas-Cao]

suggest to call general_agent.run("写一篇关于人工智能发展历程的简短文章") as well .

• general_agent : for testing default prompt of MyReflectionAgent
• code_agent : for testing customized prompt of MyReflectionAgent