Implement in-memory caching for parsed templates and include dependencies

[Matographo]

Summary

Add cache keys, cache invalidation, and correctness checks for repeated renders of unchanged templates.

Why

Caching is a first-class performance goal for both template parsing and include-heavy rendering.

Acceptance Criteria

• Warm-cache runs outperform cold-cache runs measurably
• Changing an included file invalidates dependent cached results
• Cached renders are byte-identical to cold renders

Dependencies

• Depends on Implement include processing with preprocessing and cycle detection #20
• Depends on Implement the core renderer for strings and single files #22
• Depends on Implement directory rendering with relative path preservation #23