Module scoring 

Core Scoring Algorithms for Heuristic File Prioritization

Implements the multi-dimensional scoring system originally prototyped in Python and now maintained directly inside the Scribe Rust workspace:

Scoring Formula

final_score = Σ(weight_i × normalized_score_i) + priority_boost + template_boost

Where component scores include:

• Documentation importance (doc_score)
• README prioritization (readme_score)
• Import graph centrality (import_score)
• Path depth penalty (path_score)
• Test-code relationships (test_link_score)
• Git churn recency (churn_score)
• PageRank centrality (centrality_score, V2)
• Entrypoint detection (entrypoint_score)
• Examples detection (examples_score)

Re-exports

pub use normalization::NormalizationStats;

pub use normalization::NormalizedScores;

pub use types::HeuristicWeights;

pub use types::ScoreComponents;

pub use types::ScoringFeatures;

Modules

final_scoring

Final score calculation and combination logic

normalization

Score normalization logic for consistent heuristic scoring

types

Core data types for the heuristic scoring system

Structs

HeuristicScorer

Main heuristic scorer that coordinates all scoring components

Enums

WeightPreset

Preset weight configurations for common use cases