Describe collection names that follow Arc's naming policy on the JVM and TypeScript backends

[woksin]

The camel-casing scenario told Kotlin, Java and TypeScript readers to work around a collection-name mismatch: Chronicle wrote projected read models to an unpluralized collection, and Arc read a pluralized one. That mismatch is fixed across the stack:

• Chronicle.Kotlin 6.10.0 (Let the client name read model containers with an opt-in policy Chronicle.Kotlin#113) and @cratis/chronicle 6.29.0 (Add an opt-in read model naming policy for container names Chronicle.TypeScript#152) add a read-model naming policy for container names.
• Arc.Kotlin (Name Chronicle read model collections with Arc's naming policy Arc.Kotlin#262) and Arc for TypeScript (Store Chronicle read models in the collection Arc reads them from Arc.TypeScript#144) pass their MongoDB naming policy into the Chronicle client, so projections land in the collection Arc reads.

The Arc snippets this page tabs already show the new setup. This updates the prose to match:

• Kotlin and Java: collection names need no configuration. The section says which Chronicle.Kotlin version is required, how to choose other names (an Arc NamingPolicy bean, or a Chronicle ReadModelNamingPolicy bean), and what an app that declares its own ChronicleOptions must do. The pluralizeReadModelNames = false workaround and the Update managed AI corpus to v2.0.8 #105 link are gone.
• TypeScript: collection names follow Arc for read models listed in withMongoDB's readModels; others keep their identifier. The section gives the required @cratis/chronicle version and the own-client case. It also says withMongoDB's database must match the event store name. The collectionName: type => type.name workaround, the .containerName() caveat and the Revise the capstone and vertical-slice pages for every backend #149 link are gone. It no longer says the TypeScript client has no naming policy option.

Checked locally, with the siblings taken from origin/main:

• node --test scripts/*.test.mjs, npm run lint:docs and npm run audit:variant-docs pass.
• lint-prose reports only the page's existing "camelCase" spelling flags.

Review: same-provider (Anthropic-only); I wrote the change and checked it against the merged Arc snippets.

Refs #19