TanStack
OSS
ai
Sign in / Sign up
Open main menu
ai
GitHub
Overview
Runs
Analytics
Loading workspace stats
Loading workspace insights...
Statistics interval
7 days
30 days
Latest CI Pipeline Executions
Status
Fix filter
Filter
Fuzzy
Filter range
Sort by
Sort by
Start time
Sort ascending
Sort descending
Succeeded
feat/use-chat-per-turn-schemas
876cd0e4 feat(ai-react,ai-client,ai): per-turn schemas for structured-output chat Builds on the structured-output MessagePart design from PR #577. The hook now accepts `outputSchemas: Record<string, SchemaInput>` and a per-call `sendMessage(content, { schema: 'name' })` option so a single chat surface can multiplex multiple structured artifacts (e.g. recipe vs weekly plan). Each assistant message's structured-output part carries `schemaName` as the runtime discriminator. A new `getStructuredPart(message, name)` helper on the hook return gives a typed view of the part narrowed against the matching schema in `outputSchemas`. Server-side, `chat()` accepts an optional `schemaName` that propagates through `structured-output.start` / `structured-output.complete` events. The client-side StreamProcessor stamps it onto the part automatically. New example: `/generations/structured-multi` in ts-react-chat demonstrates two schemas in one conversation with type-safe render branching. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
by Tom Beckenham
T
Succeeded
feat/use-chat-per-turn-schemas
876cd0e4 Merge 0c23c87d675be5cbb03cccde6000f5f14568a595 into 64175f85801b764e10cfe867074939f5211f623a
by Tom Beckenham
T
Previous page
Previous
Next
Next page