docs: Add Algorithm, Shadow Logic, and Discovery Tiers sections to README

[Copilot]

The README lacked any technical depth — no explanation of the scoring model, edge-case detection, or how the assessment progresses to a result.

Changes

• The Algorithm — Documents the Weighted Trait Matrix: archetypes as 5-dimensional trait vectors (Dominance, Empathy, Independence, Creativity, Discipline), with user results computed via weighted dot-product similarity:

  similarity(user, archetype) = Σ (weight_i × user_trait_i × archetype_trait_i)
  

• Shadow Logic — Describes unbalanced-profile detection. Example: high Independence + Empathy < 20% surfaces a ⚠️ Rebel Shadow warning. Covers that triggers are evaluated across all trait pairings.

• Discovery Tiers — Outlines the three-stage narrowing process:

  • Core Driver — Motivational quadrant placement (Power, Connection, Expression, Structure)
  • Action Style — Execution strategy layer; narrows candidates within the quadrant
  • Refinement — Decision-tree questions for final archetype lock + Shadow warning resolution

Original prompt

Suggested README.md Highlights

If you are setting up the repository now, including these sections will make it stand out:

The Algorithm: Briefly explain the "Weighted Trait Matrix." (e.g., “Each archetype is a vector in a 5-dimensional trait space; user results are calculated via dot-product similarity.”)

Shadow Logic: Mention how the engine detects "unbalanced" profiles (e.g., high Independence but <20% Empathy triggers the "Rebel Shadow" warning).

Discovery Tiers: 1.  Core Driver: Quadrant identification.
2.  Action Style: Execution strategy.
3.  Refinement: Final archetype lock.