A read-only registry for approved future discovery providers, access modes, terms review, quota boundaries, attribution rules, Vault ingestion boundaries, and capsule readiness gates.
Every TheoB pathway can move through Past, Present, and Future without losing context.
Read current signals, conditions, and live context.
Voice ready
Discovery Provider Registry defines where TheoB may eventually retrieve information: search APIs, encyclopedic sources, academic systems, government datasets, news feeds, cacao domain feeds, and multimodal file sources. Everything is listed. Nothing is live.
Discovery Provider Registry is ready as a non-destructive provider map. Live retrieval, credentials, provider polling, Vault ingestion, and capsule compression remain disabled until provider terms, credentials, quotas, attribution, and governance gates are approved.
Retrieve broad web results through an official Google search interface when approved.
Do not scrape Google result pages. Use official access only.Retrieve broad web results through an official Bing-compatible API when approved.
Respect quotas, billing, attribution, and caching limits.Retrieve independent web search results through approved Brave Search access.
Use as a comparison provider, not a single truth source.Provide entity context, definitions, and source trails for baseline understanding.
Treat as contextual knowledge and inspect citations for factual claims.Provide structured entity relationships for discovery, timelines, and knowledge graphs.
Validate important claims against primary or higher-authority sources.Retrieve DOI metadata, scholarly publication trails, authors, journals, and citation context.
Metadata is not the same as full paper verification.Retrieve academic works, institutions, authors, concepts, and citation relationships.
Separate scholarly metadata from claim-level truth.Retrieve paper abstracts, citation contexts, authors, and research graph signals.
Do not over-trust abstracts without paper-level review.Retrieve biomedical and life-science publication records when relevant.
Medical or health-related outputs require careful uncertainty and source boundaries.Access US public datasets for evidence-backed civic, economic, climate, and infrastructure discovery.
Dataset freshness, schema, and provenance must be checked.Retrieve climate, oceanic, atmospheric, and weather-adjacent datasets.
Use dataset metadata and temporal resolution before drawing conclusions.Retrieve space, earth observation, planetary, and scientific datasets.
Preserve mission, instrument, and dataset provenance.Retrieve global development, economic, demographic, and regional indicators.
Indicators need date, country, methodology, and revision awareness.Retrieve agriculture, food, crop, and cacao-adjacent production context.
Agricultural statistics must preserve region, crop definitions, and reporting year.Track current events across approved publishers and feeds.
Do not treat speed as accuracy. Do not store copyrighted news bodies without rights.Monitor public publisher feeds for current developments and source comparison.
Store metadata, summaries, and links only unless rights allow more.Curate cacao research, agriculture, culture, ceremony, manufacturing, and sustainability sources.
Cacao cultural and scientific sources must be separated, attributed, and respectfully contextualized.Prepare future ingestion for images, PDFs, CAD, schematics, architectural plans, maps, and datasets.
Do not process files automatically without type validation, rights awareness, redaction, and human-readable interpretation boundaries.Search and data providers must use approved APIs, feeds, datasets, or connectors.
No scraping provider result pages.Each provider needs terms, quota, caching, attribution, and storage review before live use.
Discovery cannot be built on illegal or brittle data access.The registry defines providers but does not query them.
Provider listing is not provider activation.Provider results cannot be stored in the Vault until ingestion readiness is approved.
Do not store external content before redaction, retention, and licensing gates.Provider results cannot be compressed into Intelligence Capsules until capsule readiness exists.
Do not compress away uncertainty or source trail.Source attribution must remain attached to all provider-derived records.
Discovery without attribution becomes black-box copying.Discovery providers will eventually need health, quota, and latency monitoring.
Do not mix operational provider health with discovery provider activation yet.Turning on live retrieval for any provider must require founder-approved configuration.
No accidental live search, billing, quota burn, or rights violation.