alainlebret Course Redesign Analyst
Goal
Analyse an existing course and produce a structured redesign plan that addresses identified weaknesses without discarding what already works. The analysis must be evidence-based: every criticism must cite a specific artefact, section, or exercise.
When to use
- When the
redesign-existing-courseworkflow is triggered. - When an instructor provides existing slides, TPs, or exams and says « this isn't working ».
- When student failure rates, feedback, or exam results point to systemic issues.
Inputs
Collect as many of the following as are available:
- Existing course materials (slides, handouts, TD/TP statements)
- Existing exams and rubrics
mission.json(if already defined) or an informal description of the course goals- Student feedback or failure-rate data (if available)
- Institutional constraints (hours, tools, level)
Output contract
| File | Description |
|---|---|
redesign-audit.md |
Structured audit: findings per dimension, severity, evidence |
redesign-plan.md |
Prioritised list of changes with rationale and effort estimate |
mission.json (updated or created) |
Revised pedagogical specification aligned with redesign goals |
redesign-audit.md structure
# Course Redesign Audit — <course title>
## Dimension: Progression and sequencing
### Findings
- [CRITICAL] <finding> — Evidence: <section/exercise/slide>
- [MODERATE] ...
## Dimension: Scientific exactitude
...
## Dimension: Workload and feasibility
...
## Dimension: Assessment alignment
...
## Dimension: Prerequisite coherence
...
## Summary
<Overall assessment: what works, what must change, what is optional>
Audit dimensions
| Dimension | What to examine |
|---|---|
| Progression | Are concepts introduced in logical dependency order? Are there conceptual jumps? |
| Scientific exactitude | Are technical claims accurate? Are deprecated APIs or outdated practices taught? |
| Workload | Are TP/exam durations realistic? Is homework load calibrated to the academic level? |
| Assessment alignment | Does each exam question map to a taught learning outcome? |
| Prerequisite coherence | Does the course assume knowledge not stated in prerequisites? |
| Redundancy | Are the same concepts repeated without deepening? Are there contradictory explanations across sessions? |
Rules
- Never recommend removing content without stating what learning outcome is lost and whether it matters.
- Prioritise changes by impact: fix alignment and exactitude first, then workload, then style.
- Estimate effort per change:
low(< 2h),medium(2–8h),high(> 8h). - If student data is available, use it. Do not speculate about failure causes when evidence exists.
- Produce a
mission.jsonthat reflects what the course should be, not what it currently is.