Changes in version 0.1.0 (2026-05-18)                  

  - Initial CRAN submission.
  - Complete pipeline with orm_load(), orm_run() and orm_report().
  - Added orm_run_guarded() for running the complete ORISMA pipeline
    with a relevance-control layer.
  - Added orm_relevance_guard() to flag or exclude records with weak
    topic or occupational relevance.
  - Added three relevance guard modes:
      - flag: excludes only clearly off-topic records and flags
        uncertain records for review.
      - conservative: default balanced mode; excludes clearly off-topic
        and likely non-occupational biomedical/clinical records while
        retaining uncertain records for review.
      - strict: excludes records with weak occupational context for
        rapid practitioner-oriented screening.
  - Three original bibliometric indicators: WRDI, RCS and MGP.
  - Two preventive intelligence indicators: ASS and Bridge Article
    Score.
  - Built-in occupational risk dictionary with 58 categories organised
    in six blocks: Safety, Industrial Hygiene, Ergonomics,
    Psychosociology, Biological Hazards and Emerging Technologies.
  - Bilingual outputs (EN/ES): HTML reports, occupational risk sheets,
    extraction matrices and priority rankings.
  - Automatic dimension detection via normative blocks.
  - Manual validation support with orm_validate().
  - Sample dataset orisma_sample included for examples and testing.