Skip to content

Course Description Quality Assessment

Skill: course-description-analyzer v0.03 File assessed: docs/course-description.md (version prior to rewrite)

1. Overall Score

17 / 100

2. Quality Rating

0–39: Poor — Major revision required

The original description read well as marketing copy for the course, but it was missing every structural element a learning-graph generator needs to reliably produce 200+ well-scoped concepts: no itemized topic list, no excluded-topics boundary, and no Bloom's Taxonomy outcomes at all.

3. Detailed Scoring Breakdown

Element Points Possible Points Earned Notes
Title 5 5 Clear and descriptive: "Learning Graphs: The Key to Intelligent Textbooks"
Target Audience 5 0 No audience stated anywhere in the document
Prerequisites 5 3 Technical prerequisites given (internet, browser, GenAI tool access) but no explicit statement of required background knowledge, and no "None" fallback
Main Topics Covered 10 4 Topics are mentioned in narrative prose ("Why Learning Graphs" section) but never itemized as a scannable list
Topics Excluded 5 0 No boundaries set — scope is undefined
Learning Outcomes Header 5 0 No "After this course, students will be able to..." statement
Remember 10 0 Absent
Understand 10 0 Absent
Apply 10 0 Absent
Analyze 10 0 Absent
Evaluate 10 0 Absent
Create 10 0 Absent
Descriptive Context 5 5 "Why Learning Graphs" section gives strong motivating context (AI slop problem, coherence, roadmap framing)
Total 100 17

4. Gap Analysis

  • No target audience — without this, learning-graph generation cannot calibrate concept depth/vocabulary (e.g., "professional" vs. "high school").
  • No itemized topic list — the generator needs a scannable list of 5–10 topics to seed concept-count estimation; prose paragraphs undercount.
  • No excluded topics — without a stated boundary, concept enumeration tends to sprawl into adjacent domains (e.g., general graph theory, general LMS administration).
  • All six Bloom's Taxonomy levels missing — this is the single biggest gap. Bloom's outcomes are what let the generator produce concepts spanning recall-level vocabulary through capstone-level synthesis. Without them, a generator tends to only produce "Remember"- and "Understand"-level concepts, undercutting the 200-concept target and the diversity of concept types.

5. Improvement Suggestions (Priority Order)

  1. Add an explicit Target Audience line.
  2. Add six Learning Outcomes sections (Remember, Understand, Apply, Analyze, Evaluate, Create), each with 4–5 specific, verb-led outcomes.
  3. Convert the narrative topic description into an itemized Main Topics Covered list.
  4. Add a Topics Not Covered section to bound scope.
  5. Tighten Prerequisites into a clear required-vs-helpful split.

6. Concept Generation Readiness (Original)

At 17/100, the original description could realistically seed only 40–60 concepts before a generator would need to start improvising un-scoped material. The topic breadth (CDGs, taxonomies, graph algorithms, vis.js, agents) is actually reasonable — the missing piece was outcome granularity, which is what Bloom's-level outcomes provide.

7. Next Steps

This assessment was used to drive an immediate rewrite of docs/course-description.md (same session). See the updated file for the revised score. Recommended follow-on: re-run this analyzer against the rewritten file, then proceed to the learning-graph-generator skill once the score is ≥ 85.