Questions teams ask before depending on shared schemas.

Because private data shapes create a tax on every integration. Teams keep redesigning Customer, Address, Event, Invoice, Policy, and countless other contracts, then pay for the differences in adapters, validators, UI code, docs, tests, analytics, and agent prompts.

Rusl is a commons for reusable data contracts. It gives schemas, versions, bundles, annotations, and annotation types stable places to live, be reviewed, be referenced, and be reused by humans, agents, organizations, and libraries.

Documentation helps people understand intent, but prose is a weak dependency target. Rusl makes the contract itself explicit: a schema URL can be referenced, versioned, downloaded, installed, annotated, and validated by tools.

A schema is a reusable contract for the shape and validation rules of data. Rusl currently builds on JSON Schema because it is full-featured, battle-tested, and supported across languages. Rusl is not trying to invent a new schema language; it gives schemas a registry, versions, immutable downloads, bundles, annotations, and discovery so they can be reused as shared infrastructure.

A version is an immutable published point in a schema or bundle history. Once a schema version is published, it is available as a cached immutable download, so consumers should depend on explicit versions instead of a mutable idea of latest.

Yes. Published schema versions are durable artifacts. Tools, services, build systems, command-line tools, and agents can cache a versioned download because the content for that version will not be rewritten.

A bundle is a versioned package of schemas. It declares a coherent dependency graph so consumers resolve the same set of contracts instead of assembling a schema set by memory or copy-paste.

An annotation is schema-validated semantic data attached to a subject. The community can add UI hints, retention policies, usage examples, compliance signals, library integrations, and other typed meaning without owning or forking the underlying schema.

A shared shape becomes more valuable every time people attach reliable meaning around it. Annotations are typed because their content is validated against annotation type schemas, so consumers and agents can depend on the shape of that meaning instead of scraping comments.

An annotation type is the community-published validation schema for a kind of semantic data. Consumers can ask for known annotation types and rely on the shape of the content they receive.

Schemas should define data shape. Annotations let UI, governance, audit, storage, documentation, and internal policy meanings evolve around that shape without bloating or forking the schema.

Yes. Rusl allows multiple annotations of the same type on one subject. Precedence is a consumer or annotation-type policy, not a global registry rule.

Trust comes from explicit choices: publisher identity, versions, visibility, reuse signals, watchers, favorites, annotation counts, bundle inclusion, and the annotation types a consumer understands.

The Rusl CLI is the schema package manager. A local bundle manifest controls what gets fetched, resolved, and installed into the application directory. From there, schemas can be used by the app, committed according to the project workflow, or fed into generators.

Agents need standards-based contract references for the same reason applications needed networks and APIs: private islands do not scale. Agents can use the CLI today as a concrete interface for schema workflows. A planned MCP server will let agents drive registry discovery and API workflows through Model Context Protocol.