Add Niranjan Praveen submission: Mastery Progression Engine

[Niranjan1Praveen]

This submission adds a Mastery Progression Engine to PyBe, addressing the
progression gap discussed in the mentor sessions: when is a learner ready to
advance, and what should they attempt next?

What it does

• Per-concept mastery scores (0-100) computed from existing session data
  (exposure, prompt quality, abstraction depth, reflection, misconception penalty)
• Stage gating: Explorer and Builder scenarios stay locked until the learner
  masters half of the previous stage's concepts (threshold 70, configurable)
  across at least 3 sessions
• A next-scenario recommender that surfaces the 3 unattempted scenarios with
  the highest growth potential for the learner's weakest concepts

Pedagogical grounding

• Bloom's Learning for Mastery (1968): explicit mastery thresholds instead of
  free navigation
• Piaget's stage theory: the three difficulties reframed as cognitive stages,
  explained to the learner in the UI
• Vygotsky's Zone of Proximal Development: recommendations stretch the weakest
  concepts while staying within unlocked stages

Implementation

• New: server/src/services/masteryEngine.js, server/src/routes/mastery.js
  (GET /api/mastery)
• Extended: Mastery Journey panel + difficulty locking in the client
• Deterministic and fully local (no external AI), consistent with V0
  constraints; each engine function is a clean seam for V2 LLM replacement

Submission folder: summership-26-prs/Niranjan Praveen/ (complete runnable app)