data-curation-observatory

Local-first data curation workflow using manifests, technical cards and CERTEZA/INFERENCIA/INCOGNITA reports.

Status: public-safe toolkit. Use it with synthetic or reviewed inputs first; do not point it at private folders, secrets, paid content, or unreleased IP without a separate release gate.

Problems It Helps With

• Research folder chaos: turn loose notes and files into manifests and technical cards.
• Duplicated sources: preserve the useful ficha and mark redundant source files for review instead of copying everything forever.
• Mixed certainty: separate CERTEZA, INFERENCIA, and INCOGNITA so agents stop treating every paragraph as fact.
• Publication risk: create review artifacts before deciding what can become open source, paid product, internal research, or private material.

What This Includes

• Provides generic templates for local data curation.
• Helps separate observed evidence, inference and unknowns.
• Produces review artifacts without touching input data.

What This Excludes

See PRIVATE_EXCLUSIONS.md.

License

MIT License.

Next Steps

• Add a synthetic sample folder.
• Add a manifest-generation smoke test.
• Run secret scan and path scrub before any push.