Slide Evaluate

Runs two validation agents over generated course materials and writes markdown evaluation reports.

• Program Chair — academic rigor, assessment validity, alignment with program standards, coherence.
• Test Student — clarity, engagement, learning support, accessibility.

When to invoke this skill

• User has a directory of generated slides (from course-generate or manual) and wants quality feedback
• User asks "how good is this course?" / "review these slides"
• Used as the Phase 3 evaluation in the ADDIE pipeline

Prerequisites

• pip install instructional-agents
• OPENAI_API_KEY set in the environment

Usage

python3 "${CLAUDE_PLUGIN_ROOT}/skills/slide-evaluate/scripts/evaluate.py" \
  <course_dir> [--model gpt-4o-mini] [--role both|pc|student]

Arguments

Arg	Required	Description
<course_dir>	yes	Directory with LO / syllabus / slides / scripts / assessment
--model	no	LLM model name (default: gpt-4o-mini)
--role	no	pc, student, or both (default: both)

Expected course_dir layout

course_dir/
├── learning_objectives.md
├── syllabus.md
├── assessment.md
├── chapter_1/
│   ├── slides.tex
│   ├── script.md
│   └── assessment.md
└── ...

Output

course_dir/evaluation/
├── program_chair_review.md
└── test_student_review.md

Citation

Built on the ValidationAgent from Instructional Agents (arXiv:2508.19611).