Quant Practice

Quant Practice Skill

Generate a Jupyter notebook of quant practice questions using example questions as templates. The user specifies the number of questions and topics. You generate novel questions (different tickers/values/date ranges) and provide solutions, then append them to the notebook.

Inputs to Collect

• count: number of questions
• topics: comma-separated list (e.g. returns,portfolio,signals). Empty means mixed.

Workflow

1. Create a blank notebook with a header cell:
   • Run:
     • python .claude/skills/quant-practice/scripts/generate_notebook.py --topics "<topics>" --count <count>
   • The script prints the notebook path. Capture it.
2. Select example questions from the index:
   • Run:
     • python .claude/skills/quant-practice/scripts/select_questions.py --topics "<topics>" --count <count>
   • This prints a JSON array of example question file paths.
3. Generate novel questions:
   • For each example question:
     • Read the markdown file.
     • Write a new question inspired by the example. Change tickers, numbers, or date ranges so it is materially different.
     • Write a complete solution in Python.
4. Append each question to the notebook:
   • Run:
     • python .claude/skills/quant-practice/scripts/add_question_to_notebook.py --notebook <path> --question-text "<question>" --solution-text "<solution>"
   • This appends:
     • A markdown question cell
     • An empty code cell for the user’s solution
     • A hidden solution cell

Utility

• List available topics:
  • python .claude/skills/quant-practice/scripts/get_topics.py
  • Add --only-active to show only topics from active questions.

Notes

• questions/index.json controls topic labels and which examples are active.
• If there aren’t enough active questions for a topic, either switch to mixed topics or ask the user to add more examples.
• Use simple, self-contained Python solutions (numpy/pandas/yfinance as appropriate).