Recorded from the accepted #12005 delivery (PR #13521, report: #12005 (comment)). Unassigned — recording only, for triage grading. Same rebinding class as #12038, one route over.
The gap
With HistoryMetaItemRequestSchema/HistoryMetaItemResponseSchema now declared in packages/spec/src/api/protocol.zod.ts (PR #13521), two consumers still carry pre-declaration spellings:
- The route ledger row for
GET /meta/:type/:name/history cannot yet name HistoryMetaItemResponseSchema — it predates the declaration.
client.meta.getHistory still declares its return type inline instead of binding the published response schema's inferred type.
Why it matters
Inline shapes drift silently from the contract they duplicate — the exact class #12038 is burning down. Rebinding is mechanical once the declaration exists (it now does), and each rebind removes one silent-drift surface.
Sequencing
Only after PR #13521 lands — the schemas it declares are the binding target. The dead environmentId wire-spread on the same door is already recorded on card #12005 and is NOT part of this finding.
Generated by Claude Code
Recorded from the accepted #12005 delivery (PR #13521, report: #12005 (comment)). Unassigned — recording only, for triage grading. Same rebinding class as #12038, one route over.
The gap
With
HistoryMetaItemRequestSchema/HistoryMetaItemResponseSchemanow declared inpackages/spec/src/api/protocol.zod.ts(PR #13521), two consumers still carry pre-declaration spellings:GET /meta/:type/:name/historycannot yet nameHistoryMetaItemResponseSchema— it predates the declaration.client.meta.getHistorystill declares its return type inline instead of binding the published response schema's inferred type.Why it matters
Inline shapes drift silently from the contract they duplicate — the exact class #12038 is burning down. Rebinding is mechanical once the declaration exists (it now does), and each rebind removes one silent-drift surface.
Sequencing
Only after PR #13521 lands — the schemas it declares are the binding target. The dead
environmentIdwire-spread on the same door is already recorded on card #12005 and is NOT part of this finding.Generated by Claude Code