[workshop-sim] Repair: Fix checkpoint-quality detection regex to recognize :white_check_mark: shortcode

The file to edit is .github/skills/curriculum-quantitative-assessment/curriculum_assessment.py. Locate the CHECKPOINT_RE regex pattern and update it to match both the ✅ emoji and the :white_check_mark: shortcode. Run the assessment script to verify the checkpoint_quality scores are no longer zero for the 30 main workshop steps.

Problem statement

The shared curriculum quantitative assessment rubric (.github/skills/curriculum-quantitative-assessment/curriculum_assessment.py) scores checkpoint_quality as 0.0 for all 30 main workshop steps, because CHECKPOINT_RE = re.compile(r"##\s+✅\s*Checkpoint", ...) only matches the literal ✅ emoji character. The workshop consistently uses the GitHub emoji shortcode :white_check_mark: (e.g. ## :white_check_mark: Checkpoint) instead, per this repository's own content conventions. This is a rubric-detection artifact, not a missing-checkpoint problem — checklists with 2–6 items exist at the end of nearly every step (90 files contain the white_check_mark: Checkpoint heading), but they are invisible to the shared scoring script.

Proposed change

Update CHECKPOINT_RE in curriculum_assessment.py to match both the literal ✅ character and the :white_check_mark: shortcode, for example: re.compile(r"##\s+(?:✅|:white_check_mark:)\s*Checkpoint", re.IGNORECASE). Re-run the assessment script across all 30 steps to confirm checkpoint_quality now reflects the checklist_items count already computed for each file (currently unused because has_checkpoint gates the score to 0).

Failure mode classification

Learning barrier (measurement/scaffolding-detection gap, not an actual missing scaffolding gap).

Quantitative guardrail

• Cohort mean overall_score: 6.37/10 (30 steps)
• Weakest rubric dimension across the corpus: checkpoint_quality at 0.0/10 (every step)
• Current learning KPI index: 2.9/10 (active_learning 4.17 · checkpoint_quality 0.0 · scaffolding 5.0)
• This change only corrects detection logic — it does not alter any workshop content, so overall_score for every step can only rise (as checkpoint_quality moves from 0 toward min(10, checklist_items * 2.5)), and the learning KPI index rises proportionally since checkpoint_quality carries a 2.0 weight in (2.0 × active_learning + 2.0 × checkpoint_quality + 1.5 × scaffolding) / 5.5.

Acceptance criteria

• curriculum_assessment.py's CHECKPOINT_RE matches both ✅ and :white_check_mark: checkpoint headings
• The shared curriculum quantitative assessment overall_score for affected steps (all 30 main steps) stays flat or improves after the change
• The learning KPI index (2.0 × active_learning + 2.0 × checkpoint_quality + 1.5 × scaffolding) / 5.5 stays flat or improves cohort-wide, confirming students who remain in the workshop are still building skills effectively
• No workshop checklist content is removed or weakened — this is a detection-only fix; existing checkpoints and checklist items are preserved exactly as-is

Suggested owner

copilot coding agent
Related to #3772

Generated by 🔬 Workshop Student Simulator · copilot · auto · 247 AIC · ⌖ 9.63 AIC · ⊞ 14.8K · ◷

• expires on Sep 25, 2026, 8:00 AM UTC