Goal

• Make sure that the PDF rendering of a slide .smd file is well formed

Inputs

• <SMD_FILE> lecture file to check
  • E.g., msml610/lectures_source/Lesson01.4-Brief_History_of_AI.smd

Workflow

Render the Slides

• Run

  > gen_slides.py <SMD_FILE> -- --toc_type=remove_headers
  

• Read the generated PDF <PDF_FILE>
  • E.g., msml610/lectures_pdf.tmp/Lesson01.4-Brief_History_of_AI.pdf

Check the Slide Count

• Check that the number of slides in the .smd file <SMD_FILE>

  > grep "^* " <SMD_FILE> | wc -l
  

  is the same as the number of pages in the PDF without considering the References page
• If that's not true it means that some slides * <TITLE> are rendered in more than one PDF
• Then adjust the size of the pictures and change the text so that each slide fits exactly in

Conventions

• Follow the role specified in .claude/skills/role.ai_researcher.md
• Read the conventions in .claude/skills/slides.rules.md

Verification

• The number of slides in <SMD_FILE> is the same number of slides in the <PDF_FILE> file